..

2.
.1

._

\e:

:13. ..
‘ \

43.. . 1 .

 

H LIBRARY
' Michigan State
University

This is to certify that the
thesis entitled

LIFE CYCLE ASSESSMENT AND BIODEGRADABILITY OF
BIOBASED COMPOSITES

presented by
Salil Arora

has been accepted towards fulfillment
of the requirements for the

MS. degree in Chemical Engineering and
Materials Science

I ”j "W
5% 0/2/9102; / .1

Major Professor’s Signatfite

 

12/1 5/2006

Date

MSU is an Affirmative Action/Equal Opportunity Institution

PLACE IN RETURN BOX to remove this checkout from your record.
To AVOID FINES return on or before date due.
MAY BE RECALLED with earlier due date if requested.

 

DATE DUE

DATE DUE

DATE DUE

-Ann

 

’0 NOV ”0 PW

 

 

 

 

 

 

 

 

 

 

 

 

 

 

2/05 c:/CIR ateDue.lndd-p. 15‘

 

LIFE CYCLE ASSESSMENT AND
BIODEGRADABILITY OF BIOBASED COMPOSITES

By

Salil Arora

A THESIS
Submitted to
Michigan State University
in partial fiilfillment of the requirements
for the degree of
MASTER OF SCIENCE

Department of Chemical Engineering and Materials Science

2007

ABSTRACT

LIFE CYCLE ASSESSMENT AND BIODEGRADABILITY OF BIOBASED
COMPOSITES
By
Salil Arora

The current research analyzes the sustainability of novel biobased composites
developed at the Composite Materials and Structures Center (CMSC) as an alternative
to petroleum-based composites for automotive applications. A Comparativecradle-to-
pellet life cycle assessment (LCA) of Polyhydroxybutyrate (PHB)-Kenaf and
Polypropylene (PP)—Glass composites was carried out based on ISO 14040 series of
standards. Inventory analysis for both the composites included resource consumption
and emissions, and for a better understanding of this data, CML impact assessment
methods were used to classify inventory flows based on their environmental impacts.
Impact assessment results present qualified improvement in the environment profile of
biobased composites, with significant reduction in energy consumption and global
warming potential, and increased eutrophication and acidification potential for PHB-
Kenaf composites. Finally, the sensitivity of the LCA results due to use of different
allocation and impact assessment method was analyzed. In the second part of the
research, the biodegradability of these composites under controlled composting
environment was assessed based on ASTM D5338 standard. The study, carried out for
more than two months, revealed that biobased composites degraded substantially,
evident from the quantitative evaluation of carbon dioxide emissions, as well as from the
visual inspection of the composite samples. LCA results in combination with
biodegradability assessment, confirmed the sustainability of the biobased composites

analyzed.

TABLE OF CONTENTS

LIST OF TABLES ......................................................................................................... vii
LIST OF FIGURES ......................................................................................................... x
CHAPTER 1: INTRODUCTION ...................................................................................... 1
REFERENCES: .................................................................................................................. 6
CHAPTER 2: LITERATURE REVIEW ............................................................................ 7
2.1 BIOBASED COMPOSITES .............................................................................................. 7
2.1.1 Natural fibers ................................................................................................ 10
2.1.1.1 Chemical composition ............................................................................ 11
2.1.1.2 Physical and Mechanical properties ....................................................... 13
2.1.1.3 Comparison with synthetic fibers ........................................................... 15

2.1.2 Biopolymers from polyesters ......................................................................... 16
2.1.2.1 Polyhydroxyalkanoates .......................................................................... 17
2.1.2.2 Polylactic acid ........................................................................................ 19
2.1.2.3 Polycaprolactone ................................................................................... 20
2.1.2.4 Poly(alkylene dicarboxylate) .................................................................. 21

2.1.3 PHB-Natural fiber Composites ...................................................................... 21

2.2 PREVIOUS LIFE CYCLE ASSESSMENT STUDIES ............................................................ 34
2. 2. 1 Polyhyroxyalkanoate LCA studies ................................................................ 34
2.2.1.1 Gemgross, 1999 .................................................................................... 34
2.2.1.2 Gemgross and Slater, Kurdikar et al., 2000 ........................................... 35
2.2.1.3 Akiyama et al., 2003 .............................................................................. 40

2.2.2 Biobased composites LCA studies ............................................................... 42
2.2.2.1 Wotzel et al., 1999 ................................................................................. 42
2.2.2.2 Schmidt and Beyer, 1998 ....................................................................... 44
2.2.2.3 Corbiere-Nicollier et al., 2001 ................................................................. 45

2.3 BIODEGRADATION ..................................................................................................... 48
2.3.1 AS TM standards ........................................................................................... 48
REFERENCES: ................................................................................................................ 52

CHAPTER 3: PROJECT DEFINITION .......................................................................... 54

CHAPTER 4: COMPARATIVE LIFE CYCLE ASSESSMENT OF BIOBASED

COMPOSITES AND CONVENTIONAL COMPOSITES ................................................ 55
4.1 INTRODUCTION .......................................................................................................... 55
4.2 GOAL AND SCOPE DEFINITION .................................................................................... 58
4.2.1 Goal of the study .......................................................................................... 58
4.2.2 Scope of the study ........................................................................................ 58
4.2.2.1 Functional unit ....................................................................................... 58
4.2.2.2 System boundaries ................................................................................ 61
4.2.2.2.1 PP—Glass composite system boundaries ............................................. 61
4.2.2.2.2 PHB-Kenaf composite system boundaries .......................................... 62
4.2.2.3 Allocation method .................................................................................. 64
4.2.2.4 Data quality ............................................................................................ 65
4.2.2.5 Impact assessment and Interpretation methods used ............................ 68

4.3 LIFE CYCLE INVENTORY (LCI) ANALYSIS ..................................................................... 72
4.3.1 Data categories ............................................................................................ 72
4.3.2 Glass fibers LCI ............................................................................................ 76
4.3.3 Polypropylene LCI ........................................................................................ 90
4.3.4 Kenaf fiber LCI ............................................................................................. 90

4. 3. 5 Polyhydroxybutyrate (PHB) LCI .................................................................. 104
4.3.5.1 Soybean Agriculture and Transport ...................................................... 111
4.3.5.2 Soybean Crushing ............................................................................... 129
4.3.5.3 Polyhydroxybutyrate (PHB) LCI model ................................................. 135

4. 3. 6 Composites Processing data ...................................................................... 140

4. 3. 7 Composites LCI models .............................................................................. 141

4.4 LIFE CYCLE IMPACT ASSESSMENT (LCIA) ................................................................. 150
4. 4. 1 Impact Assessment Methods ...................................................................... 150
4.4.1.1 Acidification ......................................................................................... 150
4.4.1.2 Human Toxicity, Freshwater-Aquatic and Terrestrial EcoToxicity ......... 151
4.4.1.3 Depletion of abiotic resources .............................................................. 155
4.4.1.4 Depletion of the stratospheric ozone .................................................... 156
4.4.1.5 Climate change .................................................................................... 158
4.4.1.6 Eutrophication ...................................................................................... 159
4.4.1.7 Photo-oxidant formation ....................................................................... 162
4.4.1.8 Normalization and normalization factors .............................................. 163

4. 4. 2 Impact Assessment Results ....................................................................... 166
4.4.2.1 Acidification ......................................................................................... 174
4.4.2.2 Freshwater-Aquatic EcoToxicity ........................................................... 176
4.4.2.3 Terrestrial EcoToxicity ......................................................................... 177

4.4.2.4 Human Toxicity .................................................................................... 179

4.4.2.5 Depletion of abiotic resources .............................................................. 180
4.4.2.6 Depletion of the stratospheric ozone .................................................... 182
4.4.2.7 Climate change .................................................................................... 183
4.4.2.8 Eutrophication ...................................................................................... 185
4.4.2.9 Photo-oxidant formation ....................................................................... 186
4.4.2.10 Normalization results and Discussion ................................................ 188
4.5 LIFE CYCLE INTERPRETATION .................................................................................. 194

4.5.1 Sensitivity Analysis — Effect of nitrogen allocation to Soybean cultivation 194
4.5.2 Sensitivity Analysis — Effect of using TRA CI Eutrophication impact factors. 198

4. 5.3 Analysis of Energy Consumption ................................................................ 200
REFERENCES: .............................................................................................................. 203
CHAPTER 5: BIODEGRADATION STUDY OF BIOBASED COMPOSITES .............. 206
5.1 INTRODUCTION ........................................................................................................ 206
5.2 EXPERIMENTAL SET-UP AND PROCEDURE ................................................................. 208
5.4 COMPOST CHARACTERIZATION ................................................................................ 222
5.4 COMPOST CHARACTERIZATION ................................................................................ 223
5.5 SAMPLES TESTED: PROCESSING AND CHARACTERIZATION .......................................... 227
5.5 SAMPLES TESTED: PROCESSING AND CHARACTERIZATION .......................................... 228

5.6 CONTROLLED COMPOSTING EXPERIMENT: START-UP, OPERATING PROCEDURE, AND

RESULTS ...................................................................................................................... 232
5.6.1 Start-up ...................................................................................................... 232
5. 6.2 Operating Procedure .................................................................................. 233
5. 6.3 Results ....................................................................................................... 235
5.6.3.1 Results-Compost Control ......................................................................... 237
5. 6. 3. 2 Results-Cellulose positive control ............................................................ 238
5. 6.3.3 Results-Polypropylene-Glass negative control ......................................... 240
5. 6.3.4 Results-PHB-Kenaf Composites .............................................................. 242
5. 6. 3. 5 Results-PHB-g MA-PHB-Kenaf Composites ............................................ 244
5. 6. 3. 5 Results-Neat Polymer samples ................................................................ 246
5.7 END OF STUDY OBSERVATIONS ................................................................................ 248
5. 7.1 pH determination ........................................................................................ 248
5. 7.2 Dry solids and weight loss data .................................................................. 249
5. 7.3 Sample images and thickness .................................................................... 252

5. 7.4 Plant growth test ......................................................................................... 259

5.8 DISCUSSION ........................................................................................................... 263
REFERENCES: .............................................................................................................. 269
CHAPTER 6: CONCLUSIONS AND FUTURE WORK ............................................... 271
6.1 LCA STUDY CONCLUSIONS ...................................................................................... 271
6.2 BIODEGRADATION STUDY CONCLUSIONS ................................................................... 273
6.3 FUTURE WORK AND RECOMMENDATIONS .................................................................. 274
REFERENCES: .............................................................................................................. 278

APPENDIX 4.3.4: ESTIMATED RESOURCE USE AND COSTS FOR FIELD

OPERATIONS, PER ACRE KENAF, 1999. ................................................................ 279

APPENDIX 4.3.6: ENERGY REQUIREMENT FOR EXTRUSION AND INJECTION

MOLDING ................................................................................................................... 280

APPENDIX 5.2.1: WATTAGE REQUIREMENT FOR CONTROLLED COMPOSTING

ENVIRONMENT ......................................................................................................... 286

APPENDIX 5.6.3: SAMPLE CALCULATION FOR CARBON DIOXIDE EVOLUTION ON

MASS BASIS .............................................................................................................. 291

vi

LIST OF TABLES

TABLE 2.1 CHEMICAL COMPOSITION OF NATURAL FIBERS ................................................. 12

TABLE 2.2 COMPARISON OF PHYSICAL AND MECHANICAL PROPERTIES FOR NATURAL AND
SYNTHETIC FIBERS ......................................................................................................... 14

TABLE 2.3 COMPARISON OF PROCESS AND FEEDSTOCK ENERGY REQUIREMENTS FOR SOME
BIOBASED AND FOSSIL FUEL BASED POLYMERS (GERNGROSS AND SLATER) ....................... 37

TABLE 2.4 COMPARISON OF ENERGY CONSUMPTION AND C02 EMISSIONS FOR FERMENTATIVE
PRODUCTION OF PHA USING DIFFERENT CARBON SOURCES WITH PRODUCTION OF
PETROLEUM BASED POLYMERS (AKIYAMA ET AL.) .............................................................. 41

TABLE 2.5 INVENTORY DATA FOR CRADLE TO PELLET PRODUCTION OF AUTOMOTIVE SIDE
PANEL USING HEMP FIBER-EPOXY AND ABS COPOLYMER COMPOSITES (WOTZEL ET AL.) 44

TABLE 2.6 INVENTORY AND IMPACT ASSESSMENT DATA FOR PRODUCTION OF PP—CHINA

REED FIBER AND PP—GLASS FIBER PALLET (CORBIERE-NICOLLIER ET AL.) .......................... 47
TABLE 2.7 POLLUTANT CONCENTRATIONS (FROM 40 CFR PART 503.13) ........................... 50
TABLE 4.1 AVERAGE WEIGHT OF TENSILE COUPONS .......................................................... 60
TABLE 4.2 SELECTED IMPACT ASSESSMENT CATEGORIES .................................................. 69

TABLE 4.3 SIGNIFICANT FLOWS FOR PHB-KENAF AND PP-GLASS COMPOSITES (SELECTED BY

CONTRIBUTION ANALYSIS) ............................................................................................... 74
TABLE 4.4 FIBER GLASS COMPOSITION (WEIGHT PERCENT) ............................................... 76
TABLE 4.5 TOTAL ENERGY REQUIREMENTS FOR 1 KG OF TEXTILE GLASS FIBER PRODUCTION
...................................................................................................................................... 80
TABLE 4.6 EMISSIONS FROM 1 KG OF TEXTILE GLASS FIBER PRODUCTION .......................... 81
TABLE 4.7 DEAM DATASETS USED FOR MODELING OF GLASS FIBER-MINING OF RAW
MATERIALS AND PRODUCTION PROCESS ........................................................................... 85
TABLE 4.8 LCI OF GLASS FIBER (CRADLE TO PELLET): PRODUCTION OF 1 KG OF GLASS
FIBERS ........................................................................................................................... 87
TABLE 4.9 DIESEL CONSUMPTION RATES FOR KENAF CULTIVATION AND HARVESTING ......... 91
TABLE 4.10 KENAF FIBER YIELD, FERTILIZER AND HERBICIDE APPLICATION RATES ............... 93

TABLE 4.11 EMISSION FACTORS FOR VARIOUS FARMING OPERATIONS (IN G/MJ DIESEL) ...... 94

TABLE 4.12 ENERGY CONSUMPTION VALUES PER KG OF AGROCHEMICALS PRODUCED ........ 95

vii

TABLE 4.13 DEAM DATASETS USED FOR MODELING OF KENAF FIBER CULTIVATION,
TRANSPORT, PROCESSING AND TRIFLURALIN PRODUCTION PROCESS ................................ 99

TABLE 4.14 LCI OF KENAF FIBER (CRADLE TO PELLET): PRODUCTION OF 1 KG OF KENAF
FIBERS ......................................................................................................................... 102

TABLE 4.15 SIMULATION CONDITIONS FOR POLYHYDROXYALKANOATES (PHA) PRODUCTION
USING SOYBEAN OIL AND GLUCOSE AS CARBON SOURCE (REFERENCE: AKIYAMA ET AL. 2003)
.................................................................................................................................... 109
TABLE 4.16 RAw MATERIAL USE PER KG OF PHB PRODUCED USING SOYBEAN OIL ........... 110

TABLE 4.17 EMISSION FACTORS FOR FUEL COMBUSTION IN FARMING TRACTORS (OBTAINED

FROM SHEEHAN ET AL. (1998), LCI OF BIODIESEL, NREL) .............................................. 125
TABLE 4.18 (A) SOYBEAN AGRICULTURE AVERAGE DATA (1988-2000)—INPUTS ................. 126
TABLE 4.18 (B) SOYBEAN AGRICULTURE AVERAGE DATA (1988-2000)—OUTPUTS .............. 126
TABLE 4.19 CO—PRODUCTS FROM THE SOYBEAN CRUSHING FACILITY (FROM SHEEHAN ET AL.,
1998) ........................................................................................................................... 131
TABLE 4.20 ALLOCATED INPUTS AND EMISSIONS TO SOYBEAN CRUSHING PROCESS ......... 133

TABLE 4.21 DEAM DATASETS USED FOR MODELING OF SOYBEAN AGRICULTURE AND
CRUSHING PROCESS .................................................................................................... 134

TABLE 4.22 LCI OF P(3HB-CO-5MOL°/o 3HHX) (CRADLE TO PELLET): PRODUCTION OF 1 KG
OF PHB ....................................................................................................................... 136

TABLE 4.23 COMPOSITES PROCESSING ENERGY REQUIREMENT (MJ-ELECTRICITY) PER 1000-
TENSILE COUPONS ........................................................................................................ 140

TABLE 4.24 LCI OF PP-(30WT%)GLASS COMPOSITE (CRADLE TO PELLET): PRODUCTION OF
1000 TENSILE COUPONS ............................................................................................... 142

TABLE 4.25 LCI OF PHB KENAF COMPOSITE (CRADLE TO PELLET): PRODUCTION OF 1000

TENSILE COUPONS ........................................................................................................ 147
TABLE 4.26 COMPARISON OF CML AND TRACI EUTROPHICATION POTENTIALS ................ 161
TABLE 4.27 NORMALIZATION IMPACT FACTORS FOR THE WORLD IN 1995 ......................... 165

TABLE 4.28 IMPACT ASSESSMENT RESULTS FOR PP-(3OWT%)GLASS COMPOSITE (CRADLE
To PELLET) ................................................................................................................... 168

TABLE 4.29 IMPACT ASSESSMENT RESULTS FOR PHB KENAF COMPOSITE (CRADLE TO
PELLET) ........................................................................................................................ 171

TABLE 4.30 NORMALIZED IMPACTS FOR PHB-KENAF AND PP-GLASS COMPOSITES
(REFERENCE SITUATION: WORLD, 1995) ........................................................................ 192

viii

TABLE 4.31 MODIFIED-SOYBEAN AGRICULTURE AVERAGE DATA (1988-2000), ALLOCATION OF

LEACHING BASED ON APPLICATION OF N FERTILIZER (C ALLOCATION) ............................... 195
TABLE 5.1 C02 CALIBRATION DATA ................................................................................ 221
TABLE 5.2 INTERPRETATION OF COMPOST MATURITY INDEX (ADAPTED FROM OFFICIAL

SOLVITA® GUIDELINE) ................................................................................................... 226
TABLE 5.3 MATERIALS TESTED FOR BIODEGRADABILITY .................................................. 228
TABLE 5.4 SAMPLE CHARACTERIZATION ......................................................................... 231
TABLE 5.5 CUMULATIVE AMOUNT OF CARBON PRODUCED-COMPOST CONTROL ................ 238

TABLE 5.6 CUMULATIVE AMOUNT OF CARBON PRODUCED-CELLULOSE POSITIVE CONTROL 240
TABLE 5.7 CUMULATIVE AMOUNT OF CARBON PRODUCED-PP-GLASS NEGATIVE CONTROL 241
TABLE 5.8 CUMULATIVE AMOUNT OF CARBON PRODUCED-PHB-KENAF COMPOSITES ........ 243

TABLE 5.9 CUMULATIVE AMOUNT OF CARBON PRODUCED-PHB-G MA-PHB-KENAF
COMPOSITES ................................................................................................................ 245

TABLE 5.10 CUMULATIVE AMOUNT OF CARBON PRODUCED-NEAT POLYMER SAMPLES ....... 246
TABLE 5.11 END OF STUDY PH READINGS ...................................................................... 250

TABLE 5.12 DRY SOLIDS CONTENT AND WEIGHT LOSS OF SAMPLES AT THE END OF

BIODEGRADATION STUDY ............................................................................................... 251
TABLE 5.13 END OF STUDY SAMPLE THICKNESS .............................................................. 252
TABLE 5.14 CRESS SEED-PERCENT GERMINATION RESULTS ............................................ 261
TABLE 5.15 LE'I‘I'UCE SEED-PERCENT GERMINATION RESULTS ......................................... 262

TABLE 5.16 CORRELATION BETWEEN CUMULATIVE CARBON PRODUCED/ PERCENT
BIODEGRADATION AND WEIGHT LOSS AT THE END OF THE STUDY ...................................... 268

TABLE APPENDIX 4.3.6.1 THERMODYNAMIC PROPERTIES OF THE EXTRUDED MATERIALS 281
TABLE APPENDIX 4.3.6.2 EXTRUSION ENERGY REQUIREMENT (MJ-ELECTRICITY) ............. 282

TABLE APPENDIX 4.3.6.3 INJECTION MOLDING ENERGY REQUIREMENT (MJ-ELECTRICITY). 284

LIST OF FIGURES

FIGURE 1.1 SCOPE OF NSF-PREMISE PROJECT ................................................................. 5

FIGURE 2.1(A) SCREW CONFIGURATION DURING MANUAL FIBER ADDITION; 2.1(B) SCREW
CONFIGURATION DURING AUTOMATIC FIBER ADDITION; 2.1 (c) ZSK 30 TWIN-SCREW
EXTRUDER WITH ZSB 25 SIDE FEEDER ............................................................................. 24

FIGURE 2.2 TENSILE STRENGTH AND MODULUS COMPARISON OF 30-WT% HENEQUEN FIBER
COMPOSITES (A= PHB (P—226); HENQ: HENEQUEN). (PROCESSED IN SCREW
CONFIGURATION-I (MANUAL ADDITION) AND SCREW CONFIGURATION-II (AUTOMATIC ADDITION)
...................................................................................................................................... 25

FIGURE 2.3 TENSILE STRENGTH COMPARISON OF 30-WT% NATURAL FIBER COMPOSITES (A=
PHB (P-226); HENQ: HENEQUEN; PALF: PINE APPLE LEAF FIBER) (PROCESSED IN SCREW
CONFIGURATION-II) ......................................................................................................... 29

FIGURE 2.4 FLEXURAL STRENGTH COMPARISON OF 30-WT% NATURAL FIBER COMPOSITES
(A= PHB (P-226); HENQ: HENEQUEN; PALF: PINEAPPLE LEAF FIBER). (PROCESSED IN
SCREW CONFIGURATION-ll) ............................................................................................. 30

FIGURE 2.5 IMPACT STRENGTH COMPARISON OF 30-WT% NATURAL FIBER COMPOSITES (A=
PHB (P-226); HENQ: HENEQUEN; PALF: PINEAPPLE LEAF FIBER). (PROCESSED IN SCREW
CONFIGURATION-II) ......................................................................................................... 31

FIGURE 2.6 ESEM MICROGRAPHS OF PHB+30WT°/o HEMP FIBER COMPOSITES WITH SCREW
CONFIGURATION-II .......................................................................................................... 32

FIGURE 2.7 ESEM MICROGRAPHS OF PHB+5WT°/oPHB-G-MA+3OWT% HEMP FIBER
COMPOSITES WITH SCREW CONFIGURATION-II .................................................................. 32

FIGURE 2.8 ESEM MICROGRAPHS OF PHB+30WT% KENAF FIBER COMPOSITES IN SCREW
CONFIGURATION-II .......................................................................................................... 33

FIGURE 2.9 ESEM IMAGES OF PHB+5WT°/oPHB-G-MA+3OWT% KENAF FIBER COMPOSITES
IN SCREW CONFIGURATION-II ........................................................................................... 33

FIGURE 2.10 GLOBAL WARMING POTENTIAL FOR PHA PRODUCTION IN BIOMASS AND FOSSIL
FUEL SCENARIOS AND COMPARISON WITH DIFFERENT GRADES OF PE(REPRODUCED FROM
KURDIKAR ET AL., 2000) .................................................................................................. 39

FIGURE 2.11 COMPARISON OF GLOBAL WARMING POTENTIAL FOR PHA PRODUCTION
PROCESS USING SYSTEM ALLOCATION AND MASS PARTITIONING APPROACH AND PE
PRODUTION (REPRODUCED FROM KURDIKAR ET AL., 2000) ................................................ 40
FIGURE 4.1 PHASES OF AN LCA ...................................................................................... 55

FIGURE 4.2: LIFE CYCLE ASSESSMENT STAGES AND BOUNDARIES (REF:
CSS.SNRE.UMICH.EDU) .................................................................................................... 56

FIGURE 4.3 COMPARISON OF TENSILE PROPERTIES FOR PHB - (30WT°/o)KENAF AND PP-

(3OWT%)GLASS COMPOSITES ......................................................................................... 59
FIGURE 4.4 STANDARD TENSILE COUPON FOR PHB-KENAF COMPOSITE ............................. 60
FIGURE 4.5 PP—GLASS COMPOSITES PROCESS FLOW CHART AND SYSTEM BOUNDARY ........ 66

FIGURE 4.6 PHB-KENAF COMPOSITES PROCESS FLOW CHART AND SYSTEM BOUNDARY ...... 67

FIGURE 4.7 CONNECTION BETWEEN EMISSIONS, MIDPOINT AND ENDPOINT CATEGORIES ...... 70

FIGURE 4.8 FIBER GLASS PRODUCTION [BROWN 1996, EPA 1995, EPRI 1988] ................. 77
FIGURE 4.9 MODEL FOR GLASS FIBER RAW MATERIAL MIX ................................................. 83
FIGURE 4.10 MODEL FOR GLASS FIBER PRODUCTION PROCESS ......................................... 84
FIGURE 4.11 MODEL FOR TRIFLURALIN PRODUCTION PROCESS ......................................... 96
FIGURE 4.12 MODEL FOR KENAF FIBER CULTIVATION, TRANSPORT AND PROCESSING ........ 99

FIGURE 4.13 STRUCTURE OF POLY(3-HYDROXYBUTYRATE-CO-5MOL% 3-

HYDROXYHEXANOATE) [P(3HB-Co-5MOL% 3HHx)] ........................................................ 104
FIGURE 4.14 PHA PRODUCTION PROCESS (REFERENCE: AKIYAMA ET AL. 2003) ............... 108
FIGURE 4.15 SOYBEAN CULTIVATION AND OIL MILLING DATA CALCULATED BY AKIYAMA ET AL.
.................................................................................................................................... 112
FIGURE 4.16 SELECTED IOWA REGION FOR CORN AND SOYBEAN CULTIVATION (REPRODUCED
FROM NREL/TP-510-37500) ........................................................................................ 114
FIGURE 4.17 NITROGEN TRANSFORMATION CYCLE IN A TYPICAL AGRICULTURAL SYSTEM
(REPRODUCED FROM NREL/TP-510-37500) .................................................................. 118
FIGURE 4.18 PHOSPHORUS TRANSFORMATION CYCLE IN A TYPICAL AGRICULTURAL SYSTEM
(REPRODUCED FROM NREL/TP-510-37500) .................................................................. 120
FIGURE 4.19 COMPARISON OF CALIBRATED TN MODEL WITH ACTUAL DATA AND GREET
MODEL FOR TN FLOWS To THE MISSISSIPPI RIVER (FROM NREL/TP-510-37500) ............ 122
FIGURE 4.20 COMPARISON OF CALIBRATED TP MODEL WITH ACTUAL DATA FOR TP FLOWS To
THE MISSISSIPPI RIVER (FROM NREL/TP-510-37500) .................................................... 123
FIGURE 4.21 MODEL FOR SOYBEAN CULTIVATION, HARVESTING, AND TRANSPORT ........... 128
FIGURE 4.22 OVERVIEW OF SOYBEAN CRUSHING PROCESS ............................................ 130
FIGURE 4.23 MODEL FOR SOYBEAN CRUSHING PROCESS ............................................... 132
FIGURE 4.24 MODEL FOR PHB PRODUCTION PROCESS ................................................... 136

Xi

FIGURE 4.25 MODEL FOR PRODUCTION OF PP—(30WT%)GLASS COMPOSITES .................. 142
FIGURE 4.26 MODEL FOR PRODUCTION OF PHB-(30WT%)KENAF COMPOSITES ............... 143
FIGURE 4.27 ACIDIFICATION IMPACT FOR PP-GLASS AND PHB-KENAF COMPOSITES ........ 174

FIGURE 4.28 FRESHWATER AQUATIC EcoTOXICITY IMPACT FOR PP-GLASS AND PHB-KENAF
COMPOSITES ................................................................................................................ 176

FIGURE 4.29 TERRESTRIAL EcoTOXICITY IMPACT FOR PHB-KENAF & PP-GLASS
COMPOSITES ................................................................................................................ 1 78

FIGURE 4.30 HUMAN TOXICITY IMPACTS FOR PP-GLASS AND PHB-KENAF COMPOSITES... 179

FIGURE 4.31 DEPLETION OF ABIOTIC RESOURCES FOR PP-GLASS AND PHB-KENAF
COMPOSITES ................................................................................................................ 1 81

FIGURE 4.32 DEPLETION OF THE STRATOSPHERIC OZONE IMPACT FOR PP-GLASS AND PHB-
KENAF COMPOSITES ..................................................................................................... 183

FIGURE 4.33 CLIMATE CHANGE IMPACT FOR PP-GLASS AND PHB-KENAF COMPOSITES 184
FIGURE 4.34 EUTROPHICATION IMPACT FOR PP-GLASS AND PHB-KENAF COMPOSITES 186
FIGURE 4.35 PHOTO-OXIDANT FORMATION FOR PP-GLASS AND PHB-KENAF COMPOSITEs187
FIGURE 4.36 SYSTEM-WIDE CUMULATIVE NORMALIZED IMPACTS ...................................... 193
FIGURE 4.37 NORMALIZED IMPACTS: PERCENTAGE CONTRIBUTION .................................. 193

FIGURE 4.38 EFFECT OF VARIABLE NUTRIENT ALLOCATION (FOR SOYBEAN CULTIVATION) ON
CUMULATIVE NORMALIZED IMPACTS ................................................................................ 197

FIGURE 4.39 EFFECT OF USING TRACI IMPACTS ON CUMULATIVE NORMALIZED IMPACTS... 199

FIGURE 4.40 TOTAL ENERGY CONSUMPTION FOR PP-GLASS AND PHB-KENAF COMPOSITES

.................................................................................................................................... 201
FIGURE 5.1 EXPERIMENTAL SET-UP USING CARBON-DIOXIDE TRAPPING APPARATUS ......... 210
FIGURE 5.2 EXPERIMENTAL SET-UP USING GAS CHROMATOGRAPH (REPRODUCED FROM

ASTM D 5338 - 98) ...................................................................................................... 211
FIGURE 5.3 OVERVIEW OF THE COMPOSTING STUDY - EXPERIMENTAL SET-UP ................ 214
FIGURE 5.4 MODIFIED 2 LITER ERLENMEYER FLASK ........................................................ 216
FIGURE 5.5 MODIFIED EXHAUST GAS COLLECTION SET-UP ............................................... 219
FIGURE 5.6 C02 CALIBRATION CURVE (20% - 0.2%) ....................................................... 222
FIGURE 5.7 002 CALIBRATION CURVE (3% - 0.2%) ......................................................... 222

Xii

FIGURE 5.8 CARBON DIOXIDE, AMMONIA STANDARD SCALES AND COMPOST MATURITY INDEX

TABLE (REPRODUCED FROM OFFICIAL SOLVITA GUIDELINE AND WWW.SOLVITA.COM) ........ 225
FIGURE 5.9 SOLVITA CARBON DIOXIDE TEST RESULT ..................................................... 227
FIGURE 5.10 SOLVITA AMMONIA TEST RESULT ............................................................... 227
FIGURE 5.11 002 EVOLUTION TIME PLOT FOR COMPOST CONTROL .................................. 237
FIGURE 5.12 002 EVOLUTION TIME PLOT FOR CELLULOSE POSITIVE CONTROL ................. 239

FIGURE 5.13 002 EVOLUTION TIME PLOT FOR PP-GLASS COMPOSITES NEGATIVE CONTROL
.................................................................................................................................... 241

FIGURE 5.14 002 EVOLUTION TIME PLOT FOR PHB-KENAF COMPOSITES ......................... 242

FIGURE 5.15 CO2 EVOLUTION TIME PLOT FOR PHB-G MA-PHB-KENAF COMPOSITES ....... 244

FIGURE 5.16 002 EVOLUTION TIME PLOT FOR NEAT POLYMER SAMPLES ......................... 247
FIGURE 5.17 SAMPLE 5A PP-GLASS NEGATIVE CONTROL-END OF STUDY IMAGE .............. 253
FIGURE 5.18 SAMPLE 11A PP-GLASS NEGATIVE CONTROL-END OF STUDY IMAGE ............ 253
FIGURE 5.19 SAMPLE 3A PHB-KENAF COMPOSITES-END OF STUDY IMAGE ...................... 254
FIGURE 5.20 SAMPLE 9A PHB-KENAF COMPOSITES-END OF STUDY IMAGE ...................... 255
FIGURE 5.21 SAMPLE 13A PHB-KENAF COMPOSITES-END OF STUDY IMAGE .................... 255

FIGURE 5.22 SAMPLE 4A PHB-G MA-PHB-KENAF COMPOSITES-END OF STUDY IMAGE 256
FIGURE 5.23 SAMPLE 6A PHB-G MA-PHB-KENAF COMPOSITES-END OF STUDY IMAGE 256

FIGURE 5.24 SAMPLE 10A PHB-G MA-PHB-KENAF COMPOSITES-END OF STUDY IMAGE .. 257

FIGURE 5.25 SAMPLE 12A PHB-END OF STUDY IMAGE ................................................... 258
FIGURE 5.26 SAMPLE 7B PLA-END OF STUDY IMAGE ...................................................... 258
FIGURE 5.27 SAMPLE 108 CA- (30WT%)TEC-END OF STUDY IMAGE ............................... 259

FIGURE 5.28 CYCLIC METABOLIC PATHWAY OF THE BIOSYNTHESIS AND DEGRADATION OF
P(3HB), REPRODUCED FROM ALBERTSSON, KARLSSON (1995) ...................................... 266

FIGURE 6.1 EFFECT OF FIBER WEIGHT FRACTION ON TENSILE MODULUS OF PP-KENAF FIBER
COMPOSITES (WAMBUA ET AL.3) ................................................................................... 275

FIGURE 6. 2 WEIGHT DIFFERENCE (VOLUME BASIS) FOR PHB-KENAF AND PP-GLASS
COMPOSITES ................................................................................................................ 277

Inages in this thesis are presented in color.

xiii

Chapter 1: Introduction

The concept of sustainable development was highlighted when the World
Commission on Environment and Development chaired by Dr. Brundtland,
published its report Our Common Future in April 1987. The commission aptly
defined sustainability as meeting the needs of current generation without
affecting the ability of future generations to meet their own needs. The
recommendations by the commission led to the Earth Summit-the United Nations
Conference on Environment and Development (UNCED) in Rio de Janeiro in
1992, which endorsed the Rio declaration on environment and development and
adopted Agenda 21, a 300-page plan for achieving sustainable development in
the 21’"t century. It was followed up by the World Summit on Sustainable
Development (WSSD) in Johannesburg, 2002.

There are various indicators available to highlight the importance of
sustainable development, one of the most important being: available non-
renewable energy reserves. Based on the World energy outlook-2004‘, there are
36 to 44 years of crude oil and natural gas reserves remaining as per the year
2003 consumption levels. In such a scenario the need for application of
resolution such as Agenda 21 cannot be more critical and immediate.

Since the 1992 Earth Summit, many countries (especially developed
countries) have adopted the principles of sustainable development. However,
that application has been to regulation of environmental stresses, to a lesser
extent to sustainable production, and rarely to sustainable consumption.

Phenomena such as globalisation have catalysed the economic growth in

developing countries such as China, India, etc; thus increasing the demand for
already depleting world resources‘. Therefore, the current priorities are to
develop technologies for sustainable production and improve the existing
consumption patterns.

In United States several initiatives are being taken by industries,
government, and other organizations in order to aid sustainable production and
encourage sustainable consumption. One of the recent developments is the
approval by federal government of section 9002 of the Farm Security and Rural
Investment Act of 2002 (FSRIA) in 20052. This is a significant development, since
this statute requires all the federal agencies to purchase biobased products
based on the regulations defined to implement the final rule, for all biobased
products within selected items costing over $10,000 or when the quantities of
equivalent items purchased in last fiscal year totalled $10,000 or more. This
regulation provides an incentive for new biobased products for whom the market
has yet not matured and also provides financial support for testing biobased
products in order to establish the biobased content and performance in
comparison to available products.

In my thesis research work, I have worked in the team of Michigan State
University (MSU) researchers to develop Green composites from natural fibers
and bioplastics. This research was funded by National Science Foundation (NSF)
under Product Realization and Environmental Manufacturing Innovative Systems

of Eco-Efficiency (PREMISE) program3. The objectives of this project were to

 

'An example being, China becoming the 2nd largest consumer of crude oil, overtaking Japan
consumption levels in 2004.

design and engineer, eco-friendly biobased composites for automotive
applications. Polyhydroxybutyrate (PHB) was selected as the biopolymer for this
study because it is a semi-crystalline polymer and hydrophobic. However, their
are issues with the properties and sustainability of PHB, which have prevented its
widespread use in the industry.

The main drawbacks of PHB are its brittleness, very limited processing
window, and thermal instability. In the current research project, graft
copolymerisation, maleation in particular was used to improve the physico-
chemical properties. In addition, maleated-PHB acts as a compatibilizer and has
been proved to improve biopolymer matrix-natural fibers adhesion‘. Therefore, in
this project effect of compatibilizer to improve mechanical properties of the
biocomposites was studied.

Selection of natural fibers for the project was based on the type of fiber
and the improvements related to its use. Incorporation of bast fibers is known to
improve the stiffness, while leaf fibers increase the toughness of the
biocomposites. Thus, two bast fibers, kenaf and hemp; and two leaf fibers,
henequen and pineapple leaf fiber were used in this research and the
improvement in mechanical properties due to these different fibers was studied.

To develop biocomposites, “Cascade Engineering” processing approach
was used in order to avoid multiple steps for processing of biocomposites. In this
approach PHB pellets with or without maleated-PHB (functionalised before
through reactive extrusion) and chopped natural fibers were extruded in a twin-

screw extruder ZSK 30 (Werner Pfleiderer). The extruded composite strands

were pelletized before injection molding the pellets as standard tensile coupons
for mechanical property evaluation. Details about the processing of biopolymer
and biocomposites and the optimised results have been discussed in the
Section-2a (literature review of biocomposites).

As discussed above, their are existing concerns about the sustainability of
PHBS'G'7 and therefore PHB-based composites. However, these sustainability
evaluation studies lack the detail, which is needed to reach a reasonable
conclusion. Therefore in the present NSF project, I developed the cradle to pellet
life cycle profile for PHB based on recently published data by Akiyama et al.8 In
addition life cycle data for kenaf cultivation and processing was also collected in
order to fully evaluate the sustainability of biobased composites. The overall
scope of the LCA study was cradle to pellet, i.e. evaluating from the point raw
materials were obtained to the production of finished composite.

One of the other issues with using conventional composites such as PP-
Glass, is there non-biodegradability; thus rendering disposal options of landfilling
and composting useless. Moreover, disposal by incineration is energy intensive
and known to leave toxic residues and emissions. Therefore, every disposal
option for conventional composites increases environmental burdens.

For biocomposites, both the biopolymers as well as natural fibers are
degradable in nature, but it is important to determine the rate of degradation of
composites under chosen disposal conditions and determine the effects of
disposal end products on the environment. Therefore, in the present research

biodegradability of biocomposites was determined under controlled composting

environment and the effect of disposal end product i.e. compost, on growth of
plants was determined. The overall scope of the NSF-PREMISE project is
described in Figure 1.1, with sustainability evaluation being my Master's thesis

research work.

 

 

 

Figure 1.1 Scope of NSF-Premise Project

References:

 

‘ IEA (2004), World Energy Outlook 2004, OECD/IEA, Paris

2 Guidelines for Designating Biobased Products for Federal Procurement, Section 9002,
Farm Security and Rural Investment Act of 2002 (FSRIA), 7 U.S.C. 8102. Accessed
through intemet at http://www.biobased.oce.usda.gov

3 “Design and Engineering of Green Composites from Bio-Fibers and Bacterial
Bioplastics”, NSF-PREMISE 2002 award# 0225925

‘ Add the missing reference.

5 Tillman U. Gemgross, Can biotechnology move us toward a sustainable society?
Nature Biotechnology, Vol 17, June 1999.

6 TIllman U. Gemgross and Steven C. Slater, How green are green plastics? Scientific
American, August 2000.

7 Kurdikar et al., Greenhouse gas profile of a biopolymer, Journal of Industrial Ecology, 4
(2000), 107-122.

8 M. Akiyama et al., Environmental life cycle comparison of polyhydroxyalkanoates
produced from renewable carbon resources by bacterial fermentation. Polymer
Degradation and Stability, 80 (2003), 183-194.

Chapter 2: Literature Review

In this chapter, the properties of relevant natural fibers, biopolymers, and
biobased composites are reviewed. The results of previous life cycle assessment
studies of natural fibers, biopolymers and biobased composites are also
reviewed. Finally, ASTM standards used for determining the biodegradability of

biobased composites under controlled composting conditions are reviewed.

2.1 Biobased Composites

The concept of using fibers as reinforcement for polymeric materials
originated in early 20th century, with the use of cellulose fibers in phenolics and
fiber reinforced polymeric composites were commercialized in 1940s, with glass
fibers being used as reinforcement in unsaturated polyesters”. Around the same
time in 1940, Ford Motor Company experimented with using glass fiber
reinforced soy-protein plastic composites for car panels‘. However, the use of
soy protein plastic based composites did not commercialize because of
abundance and low prices of petroleum. All of these previous examples
attempted to produce partial biobased composites, where either the polymer
matrix or fiber is non-biodegradable.

In 1989, DLR (German Aerospace Center) — Institute of Structural
Mechanics began a project to develop biobased composites from renewable
resourcesz, by using natural fibers and biodegradable polymers, so as to reduce
the reliance on non-renewable resources and alleviate the disposal problems
arising due to conventional polymeric composites, which are non-biodegradable

and difficult to recycle. Since then several review publications3,‘z,9,2 have

covered the research progress on processing natural fibers and biopolymers to
obtain biobased composites with comparable performance properties to
conventional polymeric composites. However, majority of the industrial
applications originating from this research area are partial biobased composites
(i.e. containing less than 80% renewable content). DiamlerChrysler has worked
since 1992 on developing a natural fiber reinforced components for interior and
exterior automotive applications and beginning in late 1990s, has used flax
reinforced polypropylene (PP) composites as underbody components and flax,
sisal reinforced composites as interior door panelss. In 2004, it announced using
abaca (banana fiber) reinforced PP composites for another exterior application,
as a covering for the spare wheel recess. Similar to Ford’s earlier attempts to
commercialize biocomposites in 19403, the Affordable Composites from
Renewable Resources (ACRES) research group at University of Delaware in
collaboration with John Deere 8 Co. has developed soy-based fiberglass
composites, which are used as tractor panels and hay balers‘. Natural fibers are
ductile and have superior impact resistance in comparison to glass fibers, and
thus have replaced glass fibers in components requiring better energy
absorption. For eg. since 1998, Ford Motor Co. has used PP-Kenaf composites
in interior door panel and trunk liner applications‘. Similarly, Saab (1999 Saab
98) and General Motors (2003 small passenger cars) have used LoPreFin
PP/PET/naturaI—fiber composites as full door panels“.

As evident from these industrial applications, renewable materials are

increasingly being used in composites on a partial basis, though there are few

examples of completely biobased composite applications, such as Environ®
composite board developed by Phenix Biocomposites using wastepaper and soy-
flour, and is currently marketed for home & office, and architectural non-structural
applicationsg. Their are several reasons to non-applicability of fully biobased
composites in industrial applications. First of all, higher costs of biopolymers such
as PHB, PLA in comparison to conventional commodity polymers such as PP,
LDPE, HDPE, and PVC have restricted the use of biopolymers in commercial
applicationsg. Regarding use of natural fibers in composite applications, there are
justifiable concerns related to large variation in physical and chemical properties
of natural fibers, and the resulting biocomposites. This issue, as demonstrated by
researchers at DiamlerChrysler for Green Flax, can be resolved by obtaining
fibers from a single source, limiting weather related variation by reducing the fiber
retting from 3 weeks to a maximum of 2-3 days, and standardizing the harvesting
process. Additionally, natural fibers are hydrophilic, which reduces their
compatibility with generally hydrophobic polymers and moisture
absorption/desorption can significantly reduce the mechanical properties of these
composites. Surface-chemical modification and/or coating of natural fibers are
known to improve polymer-fiber interfacial adhesion and hydrophobicity of natural
fibers and these methods are briefly reviewed in Section 2.1.1.2 of Natural
Fibers.
As discussed in the Introduction chapter of the thesis, current research
work under NSF-PREMISE program focused on developing completely biobased

composites using PHB biopolymer and various natural fibers and thus

addressing the issues highlighted above for limited commercialization of
completely biobased composites. The physical, chemical, and mechanical
properties of natural fibers relevant to the current research work are briefly
reviewed in one of the subsections below. PHB biopolymer is classified as a
polyester; and the current manufacturing process and industrial applications of
the commercially important polyesters are also reviewed below. Additionally in
the current research work, maleated-PHB was processed and used as
compatabilizer for these composites, in order to improve the fiber-matrix
interfacial adhesion. Finally, mechanical properties and morphology of these
composites was studied, so as to determine the optimal natural fiber
reinforcement and effectiveness of the compatibilizer. These results are briefly

reviewed in the subsection on PHB-Natural fiber composites.

2.1.1 Natural fibers

Natural fibers have been used since ancient times, with early applications
as textiles, ropes and more recently to produce automotive door panels,
underbody panels and dashboards, acoustic ceiling tiles, wall panels, load-
bearing composites. These fibers can be classified based on the source of origin,
as either animal or plant derived. Examples of plant-derived fibers are kenaf,
sisal, cotton; while fibers such as wool and silk are animal-derived fibers. Animal
and plant fibers can also be classified based on the difference in chemical
structure; where animal fibers are protein (polypeptide) based, and plant fibers

have cellulose as the main chemical structure. In composites industry, majority of

10

the natural fibers used are plant based, therefore only plant-derived fibers are
reviewed further.

Classification of plant—derived natural fibers is done based on the part of
the plant from which they are derived, and classified into following five
categories: 1) Bast fibers: obtained from plant stem, such as kenaf, flax, hemp,
jute, ramie; 2) Leaf fibers: examples are sisal, pineapple leaf fiber (PALF),
henequen; 3) Seed/Fruit fibers: such as cotton (from seed hair), coir (from
coconut husk); 4) Straw/Grass fibers: include wild grasses such as switchgrass,
indian grass and straw fibers from corn, wheat, and rice farming, which are
othenNise considered a waste product; 5) Wood fibers: obtained from soft and
hard woods, usually a waste from sawmills, furniture manufacture, packaging

pallets.

2.1.1.1 Chemical composition

Along with cellulose, hemicellulose and Iignin are major polymeric
components for plant-based fibers, with small amounts of pectin and wax present
for some of these fibers. Cellulose acts as a reinforcing material in plant cell
walls, with microfibrils of highly crystalline cellulose used as reinforcement in the
matrix of hemicelluloses and lignin. Chemical structure of cellulose consists of a
linear, crystalline polymer composed of 1,4-B-D-glucopyranose un'rts.

Hemicelluloses are copolymers of sugars such as glucose, mannose,
xylose, galactose, and arabinose. They cover the surface of cellulosic microfibrils
by hydrogen bonding to the surface cellulose chains. Pectins are an important

matrix component of cell walls for non-wood fibers. They are classified as

11

polysaccharides, which can have complex structures and can be branched.
Hemicelluloses and pectins are hydrophilic in nature, and are largely responsible
for the hydrophilic nature of plant-based natural fibers.

During cell development, Iignin polymer is the last component to be
incorporated in the plant cell wall, binding hemicelluloses and cellulose
microfibrils components and thus imparting rigidity, hydrophobicity and decay
resistance to the cell walls. The chemical structure of Iignin consists of a
disordered, polyaromatic, cross-linked polymer, which is obtained from the free-
radical polymerization of two or three monomers structurally related to phenyl
propane. The chemical composition of various natural fibers is presented below

in Table 2.1.

Table 2.1 Chemical Composition of Natural Fibers

 

 

Fiber Cellulose (wt%) Hemicellulose (wt%) Pectin (wt%) Lignin (wt%)
Flax 71 18.6 — 20.6 2.3 2.2
Hemp 70.2 - 74.4 17.9 — 22.4 0.9 3.7 - 5.7
Kenaf 31 -39 21.5 - 15-19
Sisal 67 - 78 10.0 - 14.2 10.0 8.0 - 11.0
PALF 70 - 82 - — 5 — 12
Henequen 77.6 4 — 8 — 13.1
Cotton 82.7 5.7 - -
Coir 36-43 0.15 -0.25 3-4 41 —45
Softwood 40 — 45 — 0 - 1 26 - 34
Hardwood 40 - 50 — 0 — 1 20 - 3O

 

As discussed above, natural fibers are composed of cellulose,
hemicelluloses, Iignin, and pectin in varying quantities (Table 2.1). Out of these
components, crystalline-cellulose mainly contributes towards the strength of

natural fibers. The natural fibers can degrade by biological action

12

(biodegradation), temperature increase (thermal degradation), and UV radiation
(ultraviolet degradation). Biological degradation of fiber components is closely
related to there moisture absorption capacity and decreases in the following
order: hemicellulose > accessible crystalline cellulose > non-crystalline cellulose
> crystalline cellulose > Iignin. In case of exposure to ultraviolet radiation,
degradation primarily happens in the lignin and to a much lesser scale in the
cellulose component. The thermal degradation of natural fiber components is in
contrast to the ultraviolet degradation, where hemicellulose and cellulose

components degrade much earlier compared to Iignin.

2.1.1.2 Physical and Mechanical properties

Natural fibers Show a large deviation in most characteristics (diameter,
length, chemical composition, crystallinity, surface properties), thus causing a
variation in mechanical properties. The deviation in fiber properties can be
attributed to the variation in quality of fibers arising due to factors such as climate
variation (during fiber cultivation), fiber maturity (at the time of harvesting),
harvesting, retting5 (water retting, dew retting or minimal retting-example is
“green” flax) and processing methods.

The strength of natural fibers has a direct dependence on crystallinity,
molecular chain orientation and is inversely proportional to the defects, cracks,
imperfections and degree of polymerization. The modulus (stiffness) of natural
fibers decreases with increase in fiber diameter. Increase in relative humidity
reduces the fiber modulus, and this effect becomes more pronounced for fibers

with higher amorphous content.

13

In addition to strength and modulus of natural fibers, fiber-matrix adhesion
is an important factor towards improving composite mechanical properties. Since
natural fibers are hydrophilic in nature and have high moisture absorption, there
compatibility with hydrophobic polymer matrixes is poor, thus causing weak
interfacial adhesion. Mohanty et al.6 have reported improvements in fiber-matrix
adhesion and composite mechanical properties by various surface-chemical
modifications of natural fibers, such as alkali treatment, etherification, acetylation,
isocyanate treatment, dewaxing, etc.

The physical and mechanical properties of various natural and synthetic
fibers have been compared previously, and are presented below for reference in

Table 2.2.

Table 2.2 Comparison of Physical and Mechanical properties for Natural and

 

 

Synthetic Fibers
Fiber Density (g/cma) Diameter (um) Tensile Young’s
strength modulus
(MPa) (GPa)
FIax7 1.4 - 1.5 19 500 — 900 50 - 70
Hemp7 1.48 25 300 - 800 30 - 60
Kenaf” 1.25 - 79.2 — 513.3 8.6 — 32.7
SisaI12 1.45 50 - 200 468 - 640 9.4 — 22.0
PALF” - 20 — 80 413 - 1627 34.5 — 82.5
Cotton7 1.5 20 300 — 600 6 — 10
Coir12 1.15 100-450 131-175 4—6
Softwood7 1.4 33 100 - 170 10 - 50
Hardwood’ 1.4 20 90 - 180 10 — 70
Glass7 2.54 10 - 20 3530 72
Aramid7 1.44 12 3600 58
Carbon7 1.75 7 3530 235

 

14

2.1.1.3 Comparison with synthetic fibers

There are several reasons for greater use of natural fibers in materials
industry, especially at the level of commodity materials, where high performance
materials are usually not required, as is the case for defense and aerospace
industry. Most of the natural fibers are renewable on annual basis, thus ensuring
an abundant and continuous supply. The degradable nature of these fibers
makes it possible to process composites, which are completely degradable (by
combining with biopolymer matrix). In contrast, composites processed from
synthetic polymers (PP, PE, PS, PVC) and fibers (glass, carbon) are non-
degradable and contribute towards ever worsening situation of properly disposing
non-degradable materials at the end of there use phase.

Based on the comparison in Table 2.2, use of natural fibers instead of
synthetic fibers in materials industry offers several advantages. These fibers in
comparison to synthetic fibers are relatively inexpensive, have low density thus a
potential for weight reduction, and better noise and thermal insulation9 (because
of there hollow tubular structure). In terms of mechanical properties, they have
comparable specific strength and high modulus. Additionally, processing natural
fiber reinforced composites poses no significant EHS (environment, health and
safety) risks (in contrast glass fiber processing can cause skin rashes and
respiratory diseases such as silicosis), and causes reduced tool wear because of
there non—abrasiveness.

Though the degradability of natural fibers is considered an advantage, it is

an undesirable attribute for many composite applications, where standard

15

outdoor performance is expected from the components during its use phase (in
years). Another drawback associated with natural fibers is their hydrophilic
nature, which reduces the compatibility with hydrophobic polymers. Both of these
issues can be resolved by surface-chemical modification and/or coating of these
fibers, as discussed before in Section 2.1.1.2. Additionally, rapid degradation of
natural fibers due to temperatures in excess of 200°C, limits the choice of
polymer matrix. Therefore, low melting polymers such as polypropylene and PHB

are selected for processing natural fiber-reinforced composites.

2.1.2 Biopolymers from polyesters

Biopolymers include the polymers, which are biodegradable and not
necessarily derived from renewable resources. Thus, biopolymers are classified
as: renewable resource based, fossil fuel based, and partial renewable resource
based polymers. Recently, biopolymers directly derived from renewable
resources or coupled with fossil fuel derived polymers have received increased
attention from scientific and industrial community, in order to reduce the reliance
on fossil fuel resources. In last 10 - 15 years, biopolymers have been

10,11,12,13,14 and

comprehensively reviewed in several journal publications
books15'16'17'18; the properties and industrial applications of these polymers are
briefly reviewed in following subsections below.

Biodegradability of polymers depends not only on the source of origin but
also on the chemical structure of the polymers”. In case of polyesters, aliphatic

polyesters are biodegradable, irrespective of the source from which they are

derived. Examples of aliphatic polyesters are PLA and PHAs, which are derived

16

from renewable resources; and PCL, PBS, which are petroleum based
biodegradable polyesters. In contrast to aliphatic polyesters, aromatic polyesters
like poly(ethylene terephthalate) (PET) are non-biodegradable. However, some
of the commercially developed aliphatic-aromatic copolyesters such as
Eastman's Eastar Bio® and BASF's Ecoflex® retain the biodegradability of
aliphatic polyesters, and have higher strength because of substitution of aliphatic
diacid monomers with more rigid aromatic diacids. The chemical structures of

these polyesters are presented below:

 

I‘ HC/CHS (EH3 C")
. 1' :01. awe/<4
n n
0 H2
Poly(lactic acid) Poly(B-hydroxybutyrate)

O 0
)0M JONOW
n n O

Poly(caprolactone) Poly(butylene. succinate)

 

 

 

2.1 .2.1 Polyhydroxyalkanoates

Polyhydroxyalkanoates (PHAS) are biodegradable polyesters, which are
commercially produced by bacterial fermentation of plant-derived sugars and oils,
and ongoing research aims to directly extract these polyesters from genetically

modified plants such as switchgrass‘g'z".

Initial discovery and chemical
identification of PHAS was done in 1920s by a French microbiologist, Maurice

Lemoigne‘", who characterized poly-3-hydroxybutyrate (PHB), a reserve polymer

17

found in several types of bacteria. However, significant development work on
PHAS did not begin until 1960s, because the bacteria which could produce these
polyesters was unknown to polymer chemists and majority of biochemistry and
microbiologists before 1958, when it was simultaneously rediscovered by
microbiologists in United States and England. lCl Zeneca started the commercial
production of PHAS in late 1980s under the trade name of Biopol®, which is a
copolyester of HB (hydroxybutyrate) and HV (hydroxyvalerate) units. This
copolyester has been used to manufacture blow molded shampoo bottles by
Wella AG and potential uses as motor oil containers, films, paper coating
material have been previously reviewed”. However, greater commercialization of
Biopol did not happen because of its higher costs compared to conventional
polymers such as PE. In 1995, Monsanto bought the process and patents
associated with production of PHAS from lCl Zeneca, and focused on PHA
production from genetically modified plants. Out of available options of corn,
sugarcane, and switchgrass, Monsanto researchers selected corn plant for
genetic modification because of its well-characterized genome. Corn plant is
genetically modified in such a way that excess PHA is expressed in the non-
edible part of the plant, i.e. corn stover (leaves and stems) and extracted using
organic solvents. Based on the sustainability assessment studies22'23'20 of PHA
production by Monsanto researchers, energy and resource consumption as well
as greenhouse gas emissions were higher for cradle-to-pellet life cycle of PHA
compared to conventional polymers such as PE. Additionally, PHA polymers

were still not cost competitive to conventional polymers. Therefore, Monsanto

18

abandoned further development work on PHAS in 1998 and Metabolix24
(Cambridge, MA) purchased its technology in 2001. As per the latest Metabolix
brochure”, “it has demonstrated economic production of PHAS by fermentation
of planbderlved sugars and oils, and is in the process of making PHAS directly in
crop plants.” In March 2006, Metabolix announced its plans in partnership with
Archer Daniels Midland Company (ADM), for annual production of 50,000 tons of
PHA natural plastics from corn starch in Clinton, Iowa adjacent to ADM’s wet
corn mill. Another company, Biomer26 (Germany) is currently pursuing
commercial scale production of PHB by bacterial fermentation using sugar

(sucrose) as a feedstock, and several grades27 of its Biomer® resin (P226, P209,

P240) are commercially available.

2.1.2.2 Polylactic acid

Polylactic acid (PLA) is a rigid thermoplastic polymer, manufactured either
by direct condensation of lactic acid or by ring-opening polymerization of the
cyclic lactide dimer. Carothers28 investigated production of PLA from lactic acid in
1932, with further development work by DuPont and Ethicon. PLA production by
direct condensation process limits the ultimate molecular weight achievable for
PLA and requires costly, environment unfriendly solvents for removal of water, in
order to obtain high molecular weight polymer. Mitsui Chemicals has developed
an improved process”, which yields high molecular weight PLA from direct
polycondensation of L-lactic acid, without using any organic solvent. In contrast
to direct condensation process, ring-opening polymerization utilizes tin catalyst to

obtain high molecular weight PLA from cyclic lactide dimer, thus eliminating the

19

need for any solvent. This production technology is used by Cargill Dow LLC,
which started full-scale production of its NatureWorksTM PLA in 2001. The
degradation of PLA after disposal takes place in high temperature and high
humidity conditions, primarily by hydrolysis, followed by microbial action‘3'18.
Because of this unique degradation characteristics, PLA based objects are stable
for years under normal conditions. Additionally, PLA is easily melt spinnable and
crystallizes upon drawing. Therefore, PLA is currently used for apparel, non-
wovens, household and industrial fabrics, carpets, fiberfill, and food packaging

applicationszg'”.

2.1.2.3 Polycaprolactone

Poly(e-caprolactone), PCL is a thermoplastic biodegradable polyester,
obtained from crude oil by chemical synthesis, and followed by ring opening
polymerization using tin catalyst. PCL biopolymer has good water, oil, solvent,
and chlorine resistance, low melting point, and good mechanical properties. The
low melting point of 60°C, limits the applications where it can be used. Therefore,
it is frequently blended with other biopolymers e.g. starch, in order to improve
their properties. Some of the common applications of PCL have been as blown
films for compost bags, as matrix in the materials ensuring controlled release of
pesticides, herbicides, and fertilizers". Additionally, PCL is used in biomedical

products14 for eg. as stiffeners for orthopedic splints, sutures, and as matrix for

bone repair.

20

2.1.2.4 Poly(alkylene dicarboxylate)

Poly(alkylene dicarboxylate) type of aliphatic polyesters are manufactured
by Showa Highpolymer, and marketed under the trade name Bionollem, which
include following grades: polybutylene succinate (PBS), poly(butylene succinate-
co-butylene adipate) (PBSA), and poly(ethylene succinate)”. These polymers
are produced by polycondensation reactions of glycols (e.g. ethylene glycol, 1,4-
butanediol) and aliphatic dicarboxylic acids (eg. succinic acid, adipic acid) and
have properties (physical and mechanical) similar to those of commodity
polymers such as LDPE. Bionolle polymers have excellent processability, and
therefore can be converted into various products such as fibers, films, bottles
using a combination of injection, extrusion and blow molding”. Unlike commodity
polymers, Bionolle polymers biodegrade in soil, compost, activated sludge,

freshwater and marine environments”.

2.1.3 PHB-Natural fiber Composites
In the previous studies on PHB based biocomposites, various natural

3"32, wheat straw”, and flax34 have been used as

fibers such as wood3°, jute
reinforcements; while determining mechanical, thermal, morphological, and
degradation properties of these composites. Additionally, some publications have
studied the effect of natural fiber reinforcement on PHB crystallinity”, and the

effect of surface/chemical modification“;32

of natural fibers on properties of PHB
based biocomposites.
The current research work under NSF-PREMISE program was comprised

of following three parts:

21

- F unctionalization of Polyhydroxyalkanoates (PHAS)

. Processing and property evaluation of PHB-Natural fiber composites

- Life cycle assessment and biodegradability evaluation of PHB-Natural
fiber composites

In the first part of the research work, solvent free functionalization of PHAS
was achieved by successfully grafting maleic anhydride (MA), octadecenyl
succinic anhydride (ODSA) on Poly(3-hydroxybutyrate) and Poly(3-
hydroxybutyrate-co-3-hydroxyvalerate) using a DSM Micro 15, twin-screw mini
extruder. The grafting of MA and ODSA was confirmed by using NMR, FTIR,
DSC, and TGA analytical techniques. The grafted PHAS were utilized as
compatibilizer in the second step of the research work, in order to improve the
fiber-matrix interfacial adhesion.

The third part of this research project involved sustainability evaluation of
PHB—Natural fiber composites based on a LCA and biodegradation study. This
part of the research work is covered in the main chapters of my thesis.

In second part of the research project, PHB-Natural fiber composites were
processed using ZSK 30 (Werner Pfleiderer) twin-screw extruder and injection
molded into tensile coupons using 85-th Cincinatti-Millacron press. The main
objectives of this part of the research work were: to study the effect of different
natural fibers, PHB-g-MA compatibilizer, different extruder screw configurations
on the mechanical properties of the biobased composites and explore ways to
commercialize the biobased composites for automotive and packaging

applications. Natural fibers selection as reinforcement was based on the

22

expected property improvement by its use, which is stiffness in case of BAST
fibers and toughness for leaf fibers. Two BAST fibers, hemp and kenaf; and two
leaf fibers, henequen and pineapple leaf fibers were used in this research. Use of
two different screw configurations was related to the goal of commercializing
processing of biobased composites. The first screw configuration had two
kneading zones (aids In polymer/fiber mixing) and three conveying zones (for
polymer transport in extruder), and processing composites in this configuration
involved manual fiber addition. In order to commercialize the processing of
biobased composites, automation of the process is very important, which in this
case was achieved by automatic fiber addition through ZSB 25 side feeder
(Coperion Corporation). For effective polymer/fiber mixing during automatic fiber
addition, extruder screw configuration was modified, and contained three
kneading zones and four conveying zones; thus enabling better mixing and
longer residence times for the polymer in the extruder. These two screw
configurations and ZSK 30 extruder set up with side feeder are presented below
in Figure 2.1 (a-c).

Injection molded tensile coupons were used to evaluate tensile, flexural,
and impact properties of PHB-Natural fiber composites. Tensile and flexural
properties were determined based on ASTM D 638 and D 790 standards using
United Calibration Corp SFM 20 machine. Impact testing of these composites
was done using Testing Machines Inc. (TMI) 43-OA-01 machine, based on

ASTM D 256 standard.

23

Fiber addition

/
IIIIIIII , \I

Fiber addition

 

 

l
I
I
l

 

 

Ill IIIIIIIIIIII II

)

b

A

 

 

 

 

 

 

(b)

 

(C)

Figure 2.1 (a) Screw configuration during manual fiber addition; 2.1 (b) Screw
configuration during automatic fiber addition; 2.1 (c) ZSK 30 twin-screw extruder
with ZSB 25 side feeder

24

Comparison of mechanical properties using screw configuration for manual and
automatic fiber addition was done by processing 30 wt% Henequen fiber
composites with both screw configurations and evaluating tensile, flexural, and
impact properties of these composites. Their was significant improvement in the
tensile modulus of the biocomposites using automatic fiber addition, which can
be attributed to better dispersion and greater mixing of fiber with polymer matrix
during extrusion process. No significant difference in other mechanical
properties: tensile strength, flexural and impact properties was noticed. A
comparison of tensile properties for PHB-Henequen fiber composites, using both

screw configurations is presented below in Figure 2.2.

lew
OTI'IW

narrower.)

 

A AM I'm ("MIDI m I'ENQ (am
addllon) addflon)

Figure 2.2 Tensile strength and modulus comparison of 30-wt% henequen fiber
composites (A= PHB (P-226); HENQ: henequen). (processed in screw
configuration-I (manual addition) and screw configuration-ll (automatic addition)

25

The effect of different natural fiber reinforcements and compatibilizer on
mechanical properties of PHB based composites was also evaluated. The tensile
properties of PHB-(30wt%) Natural fiber composites are compared below in
Figure 2.3. The use of compatibilizer slightly increases the tensile strength of
hemp, kenaf, and PALF fiber composites, while the strength of henequen fiber
composites remains the same. No significant increase in tensile modulus values
was recorded by the addition of compatibilizer. Among natural fibers, in
comparison to neat PHB samples, moderate increase in tensile strength was
recorded for all the natural fibers, with greatest increase in compatibilized hemp
and PALF fiber composite samples. The tensile modulus values increased
significantly due to natural fiber reinforcement as compared to neat PHB
samples, with greatest increase in kenaf fiber reinforced composites, whose
modulus values were also significantly higher (roughly double) compared to other
natural fiber composites.

The flexural properties of these composites (Figure 2.4), follows the same
trend as that for tensile properties, with no significant improvement in flexural
properties by use of compatibilizer, modest increase in flexural strength values,
and significant increase in flexural modulus values due to natural fiber
reinforcement. Additionally, kenaf fiber composites had significantly higher
flexural modulus values compared to other natural fiber reinforced composites.

Comparison of impact strength values of these composites is presented
below in Figure 2.5. Moderate increase in impact strength values was recorded

for hemp, PALF, and kenaf reinforced composites; while impact strength values

26

of henequen reinforced composites is significantly higher compared to neat PHB
as well as other fiber reinforced composites. The higher impact value for
henequen fiber composites was expected, since henequen is a leaf fiber and leaf
fibers are generally tougher compared to BAST fibers such as hemp and kenaf.
Use of compatibilizer had insignificant effect on impact strength of these
composites.

In addition to mechanical property evaluation, interface studies were
conducted to study the effect of compatibilizer addition on composite properties,
by investigating fiber pull-out and fiber-matrix adhesion using ESEM.

ESEM micrographs for PHB+30%Hemp and PHB+5%PHB-g-
MA+30°/oHemp composites are presented below in Figures 2.6 and 2.7. The
fracture surface of 30 wt% hemp fiber composites shows single fiber pullouts,
indicating poor fiber-matrix adhesion. In comparison compatibilized hemp fiber
composites had lesser single fiber pullouts, thus better fiber-matrix adhesion
compared to non-compatibilized hemp fiber composites. These results
corroborate the mechanical property results for hemp fiber composites, where
use of compatibilizer leads to moderate increase in tensile and flexural strength.

Similar interface studies were conducted for PHB+30%Kenaf and
PHB+5%PHB-g-MA+30%Kenaf composites, with ESEM micrographs presented
below in Figures 2.8 and 2.9. For kenaf fiber composites, addition of
compatibilizer did not improve the fiber-matrix adhesion, since fractured surfaces
had similar amount of fiber pullouts, before and after the addition of

compatibilizer. This observation is in agreement with mechanical property results,

27

since relatively small increases (compared to hemp, PALF) in tensile and flexural
strength were recorded in kenaf fiber composites due to addition of

compatibilizer.

28

(eds) snrnpow aIIsuaL

 

c.-:o=m::mc:oo 368m :_ 668823 toe... .62 26am 0:3
”3.2: 66:85: ”02w: ”ONE”: 9...: "(V mogmanoo can: .936: $31026 :OmEoQEoo 56:28 26:2. mN 2:9“.

REVERE: 544.38%: ozmssois Em: $822
-3283 u<zm§8+< -szfifii La<§8+< 56...:me 08:88} .Pm::§+< Em... $82 <

(3de mfluens eIIsuaL

3:60: :2. o
o :3:on :0... I

29

5:25.535". 320m :. commooEmv don: 52 6.88:5
SE $3.85.. ”02m... .684: 9.4. as 89.888 .8... 9:5: $.38 6 85.588 5895 .222“. SN 9:9“.

5293833.... Lin$8+<s azuxssni: . :5: 8845...
-measni u<zux$8+< -3583 8.28885 .P::5$+< 05:383.. -92.... $9... 8.2“: 382 <

              

 

  

 

      

                       

    

. _ .
O . _ . .. a... .
E I 3. 88888 A o
Emu... 5 “8.8.8.8“ a
Ambzmusw Ruin?»
.58.me \8 .8...
u :8 . 8
a.» warm» w... “x. 888%.
a»... 4.5% . 88.84.
.555 8888mm

. swam».

$8

  

it“ “9
t» t:

Rat

\NKNE
381.8 .
i . .8.
. .,. \

an?»

n
i
O
N

(9de IIIBuans Ilunxou

\\~
.. 4.x.

                     

 

a 88. .258.
w. 88%.. “8
L.» \ 2: \\ 8m .
a. 8888.... mm.
3.x 8 R a at...
xx. «.88 x =
.. , N888.

 

. Eatmek
8828.8
“384884888

odn mdn

 

 

 

 

» . r
"wwwfiflm
.4” www .

   

 

      

ad”

   

(ado) snInpow Iunxars

O
1
h
F}
I

3

 

#6 N6 0 kn

mas—v62 x2"— 0
£9.26 .3... I

 

 

30

275393356 398 :_ 69.8825 Con... “.56. 6.33:6
«:4... E885; 52m... .684: mm: as 59.888 .8: .936: $3.8 6 85.5868 5828 SEE. 8N 98.“.

 

     

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

“253.88..
<2 “3888+ <2 08.58? S: 8558+§
5.9... 8+< 422mg}. -99... 8+< 8.28383 99.58:. 029.882 $9.983 85:88}. <
t F Swizz} L .....,....H.I..W,.I.Iww,.. r a . _ . . L - c
...
a.
;.—. - m
. . R ..8 s
3” .. 88 SN _ . 3 w.
3 m
._.
‘.8 ....H." ..
8

 

.2.

31

_ :00 “m ‘\—fi1§ ‘ _ ' . _’ 200 “HR —

 

Figure 2.6 ESEM micrographs of PHB+30wt% hemp fiber composites with
screw configuration-II

v j"

— 200 um

 

Figure 2.7 ESEM micrographs of PHB+5wt%PHB-g-MA+30wt% hemp fiber
composites with screw configuration-ll

32

K_ 200 pm ‘4’ 200 um .
4-1‘ ‘ . o "-

 

Figure 2.8 ESEM micrographs of PHB+30wt% kenaf fiber composites in screw
configuration-ll

— 200 L1H] ‘— _ 200 UH] —

 

Figure 2.9 ESEM Images of PHB+5wt%PHB-g-MA+30wt% kenaf fiber
composites in screw configuration-ll

33

2.2 Previous Life lee Assessment Studies
LCA studies of natural fibers, biopolymers, and biobased composites

relevant to the current research work are briefly reviewed in this section.

2.2.1 Polyhyroxyalkanoate LCA studies

Several academic and industrial studies have analyzed the life cycle of
PHA production from biomass such as com, com stover, soybean oil by
fermentation process or direct production in genetically modified plant. The

results and conclusions of these studies are summarized below.

2.2.1.1 Gemgross, 1999

In this study, Gemgross22 has compared cradle to pellet production of
PHA with PS (polystyrene) production to obtain polymer resin. PHA is assumed
to be produced by fermentation of glucose, where glucose is obtained by com
wet milling process. Energy and raw material requirements for corn cultivation,
wet milling and PHA production process are calculated for production in US.
Based on these calculations, Gemgross has estimated that PHA fermentation
process consumes 22% more steam, 19 times more electricity, and 7 times more
water compared to conventional PS production process. Therefore, energy
requirements in terms of FFE1 (fossil fuel equivalents) for production of a
kilogram of polymer is greater for PHA (2.39 kg FFE) compared to P8 (2.26 kg
FFE), and thus PHA production has greater net greenhouse gas emissions

compared to PS production. The difference between energy requirements for

 

‘ Amount of fossil fuel (kg) consumed to produce a unit of electricity (kWh) and steam
(kg)

34

PHA and PS is not that large; however, the fact that all the energy requirement
for PHA is used as process energy, and only 1 kg of 2.26 kg is used as process
energy for PS, makes the PHA production process less sustainable.

The conclusions obtained from this study are contrary to popular
perception that biobased, biodegradable polymers are environment friendly and
sustainable. However, there are some shortcomings/limitations in the
calculations, which can significantly modify the final conclusions for this study.
First of all, fossil fuel requirements for PS production does not consider energy
required to produce/extract those fossil fuels from crude oil or natural gas, thus
lacking a true cradle to pellet scope used for other parts of this study. Secondly,
greenhouse gas emission calculations in this study does not include carbon
sequestered during corn cultivation, which significantly lowers the CO; emissions

during PHA production process.

2.2.1.2 Gemgross and Slater, Kurdikar et al., 2000

These two studies are a follow—up of the 1999 study by Gemgross and
focus on the cradle to pellet production of PHA obtained from a genetically
modified corn plant. Both the publications have a greater industrial focus
compared to the previous study by Gemgross and lists industry researchers as
co-authors.

In the first study, Gemgross and Slater23 have computed energy
requirements for PHA production from genetically modified corn plant and
compared it with PHA production from bacterial fermentation, PLA production by

bacterial fermentation, and production of PE, PET, nylon polymers derived from

35

fossil fuel resources. The comparative results in terms of process energy and
feedstock energy are presented below in Table 2.3. In addition, technical
challenges associated with PHA extraction from corn plant are also discussed.
The research work related to genetic modification of corn plant has been carried
out at Monsanto Company since 1995, when it bought the process and patents
from lCl Zeneca (see Section 2.1.2.1 for more details) and it has focused on PHA
production in corn stover, which is non-harvestable portion of corn plant. One of
the problems associated with PHA production in plants is inhibition of
photosynthesis due to high plastic content in leaves, and thus reducing grain
yields. Additionally, theoretical scale-up of the PHA extraction and collection
process concludes that the process requires large amount of solvent and
infrastructure comparable to that required for petrochemical based polymers, and
thus rightly questioning the sustainability of PHA biopolymer obtained from corn
plant. The comparison of energy requirements as shown below in Table 2.3,
further questions the “greenness” of PHA biopolymer, with process energy
required for a kilogram of PHA (from corn plant) 300 percent more than 29 MJ
required for petroleum based PE polymer. However, authors do argue that
process technologies for PHA and PLA are in the nascent stage and thus offer a
lot of scope for process improvement and energy reduction, while conventional
polymers have benefited from almost 100 years of development and innovation
in petrochemical industry. Finally, based on the energy comparisons in this study,
authors recommend using waste plant material (corn stover in current scenario)

as a source of energy during PHA production, thus replacing non-renewable

36

energy source with plant-based renewable energy source and improving the
sustainability of PHA biopolymers. The environmental impacts of this switch to
renewable energy source are discussed in their follow-up publication by Kurdikar

et al.20

Table 2.3 Comparison of process and feedstock energy requirements for some
biobased and fossil fuel based polymers (Gemgross and Slater”)

 

Process Energy (MJ/kg Feedstock Energy (MJ/kg

 

Polymer type
polymer) polymer)
PHA (from com
90 —
plant)
PHA (bacterial
81 -
fermentation)
PLA 56 -
PE 29 52
PET 37 39
Nylon 93 49

 

In the follow-up study to Gemgross and Slater”, Kurdikar et al.2°, have
compared the Global Warming Potential (GWP) for cradle to pellet production of
PHA from genetically modified corn plant with polyethylene (PE) production. The
authors have studied four different scenarios for PHA production with different
energy sources (biomass, natural gas, coal, oil) used during extraction and
compounding of PHA biopolymer. The default allocation approach, system

allocation for PHA production, allocates all the impacts for production of

37

genetically modified corn to harvested stover, while subtracting impacts for
producing same amounts of traditional corn grain. Since, corn grain yield from
genetically modified plant is less compared to traditional corn crop, this allocation
approach attributes the yield loss to harvested corn stover of the genetically
modified corn plant. In order to check the sensitivity of the results, an alternative
allocation approach, mass partitioning is used, which allocates impacts based on
the mass of harvested co-products, i.e. corn grains and stover in the current
study. The mass based comparison of GWP for different scenarios using system
allocation approach (Figure 2.10) presents a significantly lower and negative
GWP for PHA production in biomass scenario compared to fossil fuel energy
usage scenarios as well as production of different grades of PE. The lower GWP
for biomass scenario is due to carbon sequestered during corn cultivation and
002 credit obtained from excess energy produced by burning of corn stover,
which is not utilized during PHA processing. In fossil fuel scenarios, even though
carbon sequestered during corn cultivation is considered, C02 credit from
burning biomass is converted into 002 emissions from fossil fuel use in PHA
processing and stover decomposition. Since, LDPE and HDPE have lower
densities (0.93 and 0.95) compared to PHA (1.2), volumetric comparison
presents lower GWP for PE polymer grades, though it is still higher than GWP for
PHA production in biomass scenario. Sensitivity analysis using mass partitioning
approach allocates eight times the impact for farming step to harvested stover,
compared to system allocation approach, and therefore GWP for all PHA

scenarios is 20% higher in comparison to scenarios obtained using system

38

allocation approach. Despite higher C02 allocation for the farming step, biomass
scenario has negative and lower GWP compared to GWPs for different grades of
PE (Figure 2.11). Based on the results of sensitivity analysis, it is an accurate
assertion that actual GWP for PHA production lie between those calculated by
system allocation and mass partitioning approach. Therefore, PHA production
using corn stover as energy source is the only scenario with lower GWP
compared to PE production, which depends on switching from fossil fuels to corn
stover for energy requirements during processing and compounding of PHA
biopolymer. However, in their conclusions, authors have expressed doubts
regarding feasibility of such a switch, since most of the corn farming states in US

use electrical energy derived from fossil fuels.

 

 

.S
E volume-based calculations
“5
E"
:r
0
6
U
DO
Biomass, Natural Gas, Coal, Oil, LDPE resin HDPE resin
System System System System
Extension Extension Extension Extension

Figure 2.10 Global Warming Potential for PHA production in biomass and fossil
fuel scenarios and comparison with different grades of PE(reproduced from
Kurdikar et al., 2000)

39

6 000 volume-based

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

calculations
.5 4,000
g <—>
,_ 2,000 .. _. -_-_ @9fl.
0
go 0 +—-- —r- . - , fl .
E. -2,000 _——--.--
<—>

5 4.000
30 -6,000

Biomass, Biomass, Natural Gas, Natural Gas, LDPE resin HDPE resin

System Partitioning System Partitioning

Extension Extension

Figure 2.11 Comparison of Global Warming Potential for PHA production process
using System allocation and Mass partitioning approach and PE prodution
(reproduced from Kurdikar et al., 2000)

2.2.1.3 Akiyama et al., 2003

l.35 have conducted a cradle to pellet life cycle

In this study, Akiyama et a
inventory (LCI) analysis for PHA production using soybean oil and glucose as
carbon source and compared it with production of conventional polymers such as
LDPE, HDPE, PP, PS, PET on the basis of energy usage and CO2 emissions
metrics. The scope of this study is more comprehensive compared to other PHA
LCA studies reviewed in previous subsections. LCI data for agriculture and
milling of corn and soybean to obtain glucose and soybean oil respectively has
been obtained from USDA for US based farming operations. The PHA
fermentation and processing LCI data is based on a full scale simulation of PHA
production and therefore this data is more comprehensive compared to previous
studies. In the laboratory studies, authors have obtained relatively higher yields

of 0.76 g-P(3HB)/g-soybean oil compared to known high yields of 0.3 to 0.4 g-

P(3HB)/g-glucose. Therefore, in this study they have mainly focused on the PHA

40

production using soybean oil as a carbon source. In addition to energy and C02
calculations, cost estimates for PHA production are also computed. Comparison
of energy consumption and CO2 emissions for the median cost case of PHA
production using soybean oil as carbon source with PHA produced using
glucose, and petroleum based polymers is presented below in Table 2.4, and
clearly shows a lower energy consumption and CO2 emissions for PHA
production using soybean oil compared to PHA produced from glucose and

petroleum based polymers.

Table 2.4 Comparison of energy consumption and CO2 emissions for
fermentative production of PHA using different carbon sources with production of
petroleum based polymers (Akiyama et al.35)

 

 

Energy consumption CO2 emissions (kg/kg
Polymer type
(MJ/kg polymer) polymer)
PHA (soybean as C
50 0.26
source)
PHA (glucose as C
59 0.45
source)
LDPE 81 1.9
HDPE 80 1.7
PP 77 1.9
PS 87 2.6
PET 79 3.1

 

41

For the current study, LCI of PHA production was based on the median
cost scenario for PHA produced using soybean oil, because of the favorable
environmental profile and comprehensiveness of this study compared to the
previous studies by Gemgross and Slater22'23, and Kurdikar et al.2°. The results
of this study and its application in the current LCA study are discussed in detail in

Section 4.3.5.

2.2.2 Biobased composites LCA studies

A couple of LCA studies have compared the sustainability profile of
biobased composites to conventional composites, considering replacement of
glass fibers with hemp or china reed fibers. However, none of these studies have
reviewed the LCA of completely biobased composites, i.e. obtained from
biopolymers such as PHB, PLA, etc. and natural fibers. The results and

conclusions from these studies are presented below.

2.2.2.1 Wotzel et al., 1999

This study36 computes cradle to pellet inventory, impacts, and valuation of
these impacts for the production of side panel for AUDI A3 car by injection
molding of ABS copolymer resin and comparing the results with an alternative
hemp fiber (66 vol%)-epoxy resin biocomposites. The cultivation, harvesting, and
processing of hemp fibers as well as production of reinforced side panels have
been modeled to be produced in South-West Germany. Inventory data for
production of ABS copolymer and Epoxy resin is based on the data published by
APME (Association of Plastics Manufacturers in Europe). Impact assessment of

the inventory results follows UBA methods (ecological protection agency of

42

Germany), while Eco-indicator 95 method developed by Pre’ Consultants is used
to calculate Eco-indicator value from impact assessment results.

The emissions, which contribute towards majority of impacts, as well as
total energy usage for both the systems are compared below in Table 2.5. Total
energy consumption for hemp fiber-epoxy composites is 59 MJ lower compared
to ABS copolymer side panel, mainly due to low energy consumption of 5% (of
total energy used for hemp-epoxy composites) for hemp cultivation and
processing. ABS copolymer side panel has higher global warming, acidification,
winter smog, and summer smog impacts compared to hemp fiber -epoxy side
panels, primarily due to higher C02, CH4, 802 emissions from greater fossil fuel
consumption. Eutrophication and ozone depletion impacts are higher for hemp
fiber-epoxy composites because of fertilizer consumption for hemp cultivation
and fluorohydrocarbon emissions from epoxy resin production. The cumulative
impacts computed using Eco-indicator 95 valuation method are 8% lower for
hemp fiber-epoxy side panels, with majority contribution from acidification
impacts (47% for ABS copolymer and 45% for hemp-epoxy side panels).

The inventory analysis, impact assessment, and valuation results clearly
shows a superior environmental profile for production of hemp fiber-epoxy side
panels. Since the study does not consider use-phase and disposal phase of side
panels, authors have partially addressed this issue in the scenario analysis by
computing energy benefits from use phase of 200,000 km (14 yrs) and disposal
by incineration. Even though, energy benefit from incineration of ABS copolymer

is higher compared to hemp fiber-epoxy side panel (because of higher

43

hydrocarbon content for ABS copolymer), this benefit is offset by fuel savings
from using lower weight hemp fiber-epoxy side panel and presents significantly
lower overall energy consumption for hemp fiber-epoxy side panel, compared to

ABS copolymer side panel.

Table 2.5 Inventory data for cradle to pellet production of automotive side panel
using Hemp fiber-Epoxy and ABS copolymer composites (Wotzel et al.35)

 

Hemp fiber-Epoxy

 

Inventory results2 ABS copolymer
composites

Weight of side panel (9) 820 1125
Total energy usage (MJ) 73 132

(a) C02 (kg) 4.19 4.97

(a) CH4 (9) 16.96 17.43

(a) NOx (g) 18.64 14.14

(a) 802(9) 10.70 17.54

(a) Dust (g) 6.96 4.25

(w) N03' (9) 12.05 0.08 x 10‘3

 

2.2.2.2 Schmidt and Beyer, 1998

The main goal of this LCA study37 is to determine environmental benefits
of substituting glass fibers with hemp fibers as reinforcement in an insulation
component for a Ford automobile. Authors have carried out a simplified LCA with
cradle to grave scope for production of PP-EPDM (ethylene propylene diene

copolymer)-glass fiber (43 wt%) and PP-EPDM-hemp fiber (30 wt%) composites,

 

2 Prefix (a) denotes air emissions, and (w) denotes water emissions

44

assuming a use-phase of 100,000 km and equal amounts of disposal by
landfilling and incineration. For inventory analysis, data for different components
of the system are obtained from various public databases; APME for PP and
EPDM data, Pre’ and ETH databases for glass fiber data, IDEA, ETH, and
BUWAL databases for utility and other chemical datasets, and from a hemp fiber
supplier, Hemcore Limited for data regarding production of hemp. Since the
study has been conducted for internal reporting in Ford, no explicit data has been
reported, instead selected inputs and outputs are reported in terms of net
benefits from the substitution of glass fiber with hemp fiber in the insulation
component, and presenting net benefits of: 88. 90 MJ for total energy
consumption, 2.12 kg oil equivalents, 8.18 kg C02 equivalents, 0.056 kg 802
equivalents, 0.018 kg N03' equivalents. In addition to inventory analysis and
impact assessment, interpretation of the results using scenario and sensitivity
analysis is conducted, which further affirms the environmental benefits of

substituting glass fibers with hemp fibers in the insulation component.

2.2.2.3 Corbiere-Nicollier et al., 2001

Similar to the scope of the previous LCA studies of biobased composites,
this study38 compares the environmental performance of PP-China reed fiber
(CR) and PP-Glass fiber (GF) composites. A transport pallet with same tensile
modulus or stiffness for both materials is selected as a functional unit; while
scope of the study is cradle to grave, which includes cultivation, harvesting, and
processing of CR fibers, PP and glass fiber production, use as a transport pallet

for 5 years (with transport of 1000 km per year), and disposal by incineration or

45

bioactive discharge. The inventory analysis and comparison of the systems
shows lower non-renewable energy consumption, and in general lower air and
water emissions for CR pallets. CR pallets have higher soil emissions of heavy
metals, N20 and NH3 air emissions, and phosphate and nitrate water emissions,
all attributed to CR fiber cultivation and fertilizer use. A detailed analysis of these
results has been done using CST95 (Critical Surface-Time) impact assessment
method, and compared with impact assessment results from CML 92, Ecopoints,
and Eco-Indicator 95 methods. Comparison of important inventory and impact
assessment data for CR and GF pallets is presented below in Table 2.6. The
impact assessment results computed using CST95 method are comparable to
those obtained using other methods, except eutrophication and CR pallets have
lower impacts than GF pallets. In case of eutrophication, other methods include
NOx emissions as a contributor to eutrophication, while CST95 assumes
phosphate emissions as the only contributor to eutrophication (because of the
fact that European lakes are generally P-limited), and therefore CST95
eutrophication impacts are higher for CR pallets compared to GF pallets. A
scenario analysis of the end of life disposal options recommends incineration
over bioactive discharge, even though incineration of PP component of the pallet
leads to heavy metal air emissions, the energy credit obtained from incineration
and water emissions from bioactive discharge makes incineration a better
disposal solution. Finally, authors have determined the sensitivity of LCA results
due to variation in pallet’s use-phase life time, fiber composition, and use-phase

transport distances. The results of the sensitivity analysis concludes that CR

46

pallets have superior environmental profile and lower energy use, as long as the
pallets are used for a minimum of 3 years. Since the use of china reed fibers
reduce the pallet weight, whereas glass fibers increase the pallet weight (density
of glass fibers is three times the PP), the environmental benefits from china reed
fiber use become more significant due to increased fiber content in pallets and

greater transport distances during use-phase of the pallets.

Table 2.6 Inventory and Impact Assessment data for production of PP-China
Reed fiber and PP-Glass fiber pallet (Corbiere-NicoIlier et al.38)

 

 

Inventory results PP-CR pallets PP-GF pallets
Energy consumption (MJ) 717 1400
(a) C02 (kg) 42 73.1
(a) N20 (9) 2.2 1.96
(a) NH3 (9) 11.3 0.12
(a) N0x (g) 349 513
(w) Nitrate (g) 153 1.72
(w) Phosphate (g) 1.67 0.59
CML-Human Toxicity (kg 1,4- oce3 eq.) 9.04 21.2
CML-Terrestrial Ecotoxicity (kg 1,4- DCB eq.) 4480 5250
CML-Aquatic Ecotoxicity (kg 1,4- DCB eq.) 0.67 1.09
CML-Greenhouse effect (kg CO2 eq.) 40.4 75.3
CML-Ozone formation (kg ethylene eq.) 0.13 0.21
CML-Acidification (kg 802 eq.) 0.43 0.65
CML-Eutrophication (kg P043“ eq.) 0.063 0.068
CST95-Eutrophication (kg P043” eq.) 1.81 x103 8.23x10“

 

 

3 DCB = dichlorobenzene

47

2.3 Biodegradation

The standards relevant to the current biodegradation study are reviewed

briefly in the following subsection below.

2.3.1 ASTM standards

Various national and international bodies such as The American Society
for Testing and Materials (ASTM), European Committee for Standardization
(CEN), Deutsches lnstitut fiir Normung (DIN), Japanese Industrial Standards
(JIS), and International Standards Organization (ISO) have established
standards to quantify and evaluate the biodegradability of polymers in different
environments such as compost, soil, marine-waters, wastewater treatment
facilities, and aerobic digesters.

ASTM has published a series of standards for determining the
compostability of plastics, the latest being D 6400 (ASTM, 1999) “Standard
specification for compostable plastics”. This standard lays down the detailed
requirements for labeling a plastic as compostable. The other standards in this
series are as follows:

1. D 5338 (ASTM, 1998): “Standard test method for determining aerobic
biodegradation of plastic materials under controlled composting conditions”.

2. D 6002 (ASTM, 1996): “Standard guide for assessing the compostability of
environmentally degradable plastics”.

ASTM D 6400 and D 6002 have no equivalent ISO standards; however, D

5338 has an equivalent ISO standard: ISO 14855, “Determination of the ultimate

48

aerobic biodegradability and disintegration of plastic materials under controlled

composting conditions — Method by analysis of evolved carbon dioxide”.

The requirements based on D 6400 for labeling a product as compostable

can be summarized as follows:

1. Disintegration during composting - At the end of a composting study the
residual plastic material should be indistinguishable from the other organic
material in the final product. On sieving the final product, no more than 10%
of the plastic material should remain on a 2.0-mm sieve.

2. Inherent Biodegradation — A plastic material must demonstrate one of the
following rates of biodegradation (ratio of conversion of organic carbon to
carbon dioxide):

a. For a product containing a single polymer (homopolymers or random
copolymers), the biodegradation rate must be 60% as compared to a
known reference material.

b. If a product contains more than one polymer (block copolymers,
segmented copolymers, blends, or low molecular weight additives),
90% of the material should degrade in comparison to a known
reference material.

c. Products containing more than one polymer, any of which exceeds 1%
concentration, must achieve the 60% rate of biodegradation.

The standard D 6400 recommends a test period of no more than 180 days

for materials that are not radiolabeled and a test period of 365 days for

49

radiolabeled materials. Materials are radiolabeled with the (radioactive isotope)

carbon-14 if the material biodegradation rate is slow.

3. No adverse impacts on ability of compost to support plant growth — the
materials tested should not have harmful impacts on the ability of compost to
support plant growth in comparison to compost using cellulose as a control,
when the finished compost is introduced in soil. In addition, the composted
materials should not release unacceptable levels of heavy metals and other
toxic substances into the environment. The tests and guidelines to fulfill these
requirements are:

a. The composted material should have heavy metal concentrations less
than 50% of those recommended in 40 CF R part 503.13; the individual
pollutant concentrations recommended are provided in Table 2.7 (from

40 CFR part 503.13).

Table 2.7 Pollutant Concentrations (from 40 CFR part 503.13)

 

 

Pollutant Monthly average concentration (mg/kg)4
Arsenic 41

Cadmium 39

C0pper 1500

Lead 300

Mercury 17

Nickel 420

Selenium ' 100

Zinc 2800

 

b. The composted material should fulfill the requirements for plant

germination and plant growth using the cress seed test and the plant

 

4 Dry weight basis

50

growth test respectively, following OECD (Organization for Economic

Development) Guideline 208.

51

References:

 

1 Valigra L. 2000 January 20. Tough as Soybeans. The Christian Science Monitor.

2 Nickel J, Riedel U. 2003. Activities in biocomposites. Materials Today 6(4):44-48.

3 Saheb DN, Jog JP. 1999. Natural fiber polymer composites: A review. Advances in
Polymer Technology 18(4):351-363.

‘ Sherman LM. 1999 October. Natural Fibers: The New Fashion In Automotive Plastics.
Plastics Technology262-68.

5 DaimlerChrysler HighTech Report, 1999. Grown to Fit the Part. 82—85.

6 Mohanty AK, Misra M, Drzal LT. 2001. Surface modifications of natural fibers and
performance of the resulting biocomposites: An overview. Composite Interfaces
8(5):313-343.

7 Lilholt H, Mark Lawther J. 2000. Natural Organic Fibers. In: Chou T-W, editor.
Comprehensive composite materials. 1st ed. Amsterdam ; New York: Elsevier. p 303-
325.

° Arora S. 2003. Single fiber testing of Kenaf fibers. Unpublished work, Composite
Materials and Structures Center, Michigan State University.

9 Netravali A N, Chabba S. 2003. Composites get greener. Materials Today (April 2003)
22-29.

1° Albertsson AC, Karlsson S. 1995. Degradable Polymers for the Future. Acta
Polymerica 46(2): 1 14-123.

1‘ Amass W, Amass A, Tighe B. 1998. A review of biodegradable polymers: Uses,
current developments in the synthesis and characterization of biodegradable polyesters,
blends of biodegradable polymers and recent advances in biodegradation studies.
Polymer International 47(2):89-144.

‘2 Mohanty AK, Misra M, Hinrichsen G. 2000. Biofibres, biodegradable polymers and
biocomposites: An overview. Macromolecular Materials and Engineering 276(3—4):1-24.
‘3 Gross RA, Kalra B. 2002. Biodegradable polymers for the environment. Science
297(5582):803-807.

‘4 Flieger M, Kantorova M, Prell A, Rezanka T, Votruba J. 2003. Biodegradable plastics
from renewable sources. Folia Microbiologica 48(1):27-44.

‘5 Griffin GJL, editor. 1994. Chemistry and technology of biodegradable polymers. 1st
ed. London ; New York: Blackie Academic & Professional. xiii, 154 p.

1" Kaplan D, editor. 1998. Biopolymers from renewable resources. Berlin ; New York:
Springer. xviii, 417 p.

‘7 Stevens ES. 2003. Environmentally Degradable Plastics. In: Kroschwitz JI, editor.
Encyclopedia of polymer science and technology. 3rd ed. Hoboken, NJ: John Wiley. p
135-162.

‘8 Mohanty AK, Misra M, Drzal LT, editors. 2005. Natural fibers, biopolymers, and
biocomposites. Boca Raton, FL: Taylor 8. Francis/CRC Press. 875 p.

‘9 Biodegradable Performance Polymers from Switchgrass — Fact Sheet, Office of
Industrial Technologies Industries of the Future-Agriculture Program, 2002.
Downloaded from: httpzlldevafdc.nreI.gov/pdfs/8444.pdf

2° Kurdikar et al., Greenhouse gas profile of a biopolymer, Journal of Industrial Ecology,
4 (2000), 107-122.

2‘ Lenz RW, Marchessault RH. 2005. Bacterial polyesters: Biosynthesis, biodegradable
plastics and biotechnology. Biomacromolecules 6(1):1-8.

52

 

22 Gemgross TU. 1999. 'Can biotechnology move us toward a sustainable society?
Nature Biotechnology 17(6):541-544.

23 Gemgross TU, Slater SC. 2000. How green are green plastics? Scientific American
283(2):36-41 .

2‘ Boswell C. 2001 August 20 - 27. Bioplastics Aren’t the Stretch They Once Seemed.
Chemical Market Reporter.

25 http://www.metabolix.com/, Metabolix Brochure, accessed May 2006.

26 http://www.biomer.de/, Production of PHB, accessed May 2006.

27 httpr/Iwww.biomer.de/MechQatE.html

2" Carothers W.H., Dorough, G.L., van Natta, F.J. 1932. Studies of Polymerization and
Ring Formation. X. The Reversible Polymerization of Six-Membered Cyclic Esters.
Journal of the American Chemical Society 54(2):761 - 772.

29 Vink ETH, Rabago KR, Glassner DA, Gruber PR. 2003. Applications of life cycle
assessment to NatureWorks (TM) polylactide (PLA) production. Polymer Degradation
and Stability 80(3):403-419.

3° Veronika E. Reinsch SSK. 1997. Crystallization of poly(hydroxybutrate-co-
hydroxyvalerate) in wood fiber-reinforced composites. J Appl Polym Sci 64(9):1785-
1796.

3‘ Mohanty AK, Khan MA, Hinrichsen G. 2000. Surface modification of jute and its
influence on performance of biodegradable jute-fabric/Biopol composites. Composites
Science and Technology 60(7):1115-1124.

32 Mohanty AK, Khan MA, Sahoo S, Hinrichsen G. 2000. Effect of chemical modification
on the performance of biodegradable jute yam-Biopol composites. Journal of Materials
Science 35(10):2589-2595.

33 Avella M, Rota GL, Martuscelli E, Raimo M, Sadocco P, Elegir G, Riva R. 2000.
Poly(3-hydroxybutyrate-co-3-hydroxyvalerate) and wheat straw fibre composites:
thermal, mechanical properties and biodegradation behaviour. Journal of Materials
Science 35(4):829-836.

3‘ Wong S, Shanks R, Hodzic A. 2002. Properties of poly(3—hydroxybutyric acid)
composites with flax fibres modified by plasticiser absorption. Macromolecular Materials
and Engineering 287(10):647-655.

35 Akiyama M, Tsuge T, Doi Y. 2003. Environmental life cycle comparison of
polyhydroxyalkanoates produced from renewable carbon resources by bacterial
fermentation. Polymer Degradation and Stability 80(1):183-194.

3" Wotzel K, Wirth R, Flake M. 1999. Life cycle studies on hemp fibre reinforced
components and ABS for automotive parts. Angewandte Makromolekulare Chemie
272:121-127.

37 Schmidt WP, Beyer HM. 1998. Life cycle study on a natural fiber reinforced
component. SAE Total Life-cycle Conference. Graz, Austria: SAE.

3" Corbiere-Nicollier T, Gfeller-Laban B, Lundquist L, Leterrier Y, Manson JAE, Jolliet O.
2001. Life cycle assessment of biofibres replacing glass fibres as reinforcement in
plastics. Resources Conservation and Recycling 33(4):267-287.

53

Chapter 3: Project Definition

The current project is part of a research funded by National Science
Foundation (NSF) under Product Realization and Environmental Manufacturing
Innovative Systems of Eco-Efficiency (PREMISE) program. The overall
objectives of the NSF-PREMISE project were to design and engineer eco—friendly
biobased composites for automotive applications. These composites were
processed using polyhydroxybutyrate (PHB) biopolymer reinforced with one of
the BAST fibers, kenaf and hemp; or leaf fibers, henequen and pineapple leaf
fiben

The objective of the current research was to evaluate the sustainability of
these biocomposites compared to conventional PP-Glass composites. This
evaluation was done by conducting a life cycle assessment (LCA) study and by
determining the biodegradation of these composites under controlled composting
conditions. The LCA study considered cradle to pellet production of equal
volumes of PHB-(30wt%)Kenaf and PP-(30wt%)Glass composites, and
compares energy consumption and environmental impacts for the complete life
cycle of these composites, as defined by ISO 14040 series of standards. The
biodegradation of these composites was tested under controlled composting
conditions, as specified by ASTM D 5338 standard, and compares percentage

biodegradation as well as the effect of disposal end-products on growth of plants.

54

Chapter 4: Comparative Life Cycle Assessment of Biobased
Composites and Conventional Composites

4.1 Introduction

The United Nations Bruntland Commission defined sustainability as
meeting the needs of the present without comprising the ability of future
generations to meet their own needs‘. Life Cycle Assessment (LCA) is one of the
quantitative methods available to evaluate the sustainability of a process/product.

ISO 14040(1997) defines life cycle assessment as compilation and
evaluation of the inputs, outputs and the potential environmental impacts of a
product system throughout its life cycle. Figure 4.1 highlights the different phases

of an LCA, as defined by ISO 14043.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Goal and Interpretation phase Direct Applications
Scope
-product
definition developmentflt
Identlflcatlon Evaluatlon by limprovement
of slgnlflcant sorrplem check strategic planning
I to Issues _ .. . M
nven "l 58' S'M’ -public policy making
analysls musty ched<
-marketing
Impact Conduslons,
recommendations and
Assessment reportlng

 

 

 

 

 

 

 

 

 

 

 

 

(ISO 14043, 2000)

Figure 4.1 Phases of an LCA

55

Inventory analysis (as defined by ISO 14040) involves data collection and
calculation procedures to quantify relevant inputs and outputs of a product

system (Figure 4.2).

Product Life Cycle

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

M, E M, E M, E M, E ME M,.E
; 5, 3. 6 6 6
Raw Material Manufacture Use 3, Retirement Treatment
Matenal + Processing T’ &Assembly "' Service + & Recovery + Disposal
Acqursrtro'n - _ _

 

 

 

 

 

 

 

 

i v i i i i
W W W W W W
reuse

remanufacture

 

 

 

 

 

 

 

 

 

closed-loop recycle s
open-loop recycle

M, E material and energy inputs for process and distribution
W waste(gaseous, liquid, solid) output from product, process and distribution
——> material flow of product component

Figure 4.2: Life cycle assessment stages and boundaries (ref:
css.snre.umich.edu)

Fiber reinforced polymer composites have been used in various industries
for structural applications, with initial use in aerospace industry; currently they are
being used for automotive parts, building materials, sporting goods, electronic
components, etc2'3. Traditionally composites have been manufactured using
unsaturated polyesters, polyurethanes, phenolic, or epoxy resins reinforced with
glass, carbon or aramid fibers. However, the production of these conventional
polymers and fibers is energy and resource intensive as highlighted by previous
LCA studies done by the Association of Plastics Manufacturers in Europe

(APME)4, and the Department of Energy (DOE)9 and uses significant amount of

56

petroleum based materials (both as feedstock and fuel) in the production

process. Moreover, conventional composites are usually made using

nondegradable matrices and fibers, which pose significant issues in the disposal
phase because of the limited landfill capacity and also because incineration is
energy intensive and known to result in toxic residues and emissions.

Therefore, the substitution of conventional polymers with biobased
polymers and conventional fibers with natural fibers offers an instinctive solution
to the issues of reducing the dependence on petroleum feedstocks and
nondegradable composites. However, such a solution cannot be justified without
quantifying the material and energy requirements over the entire life cycle of
biocomposites, and thus addressing the following questions regarding the net-
environmental benefits associated with biocomposites:

- What is the total energy consumption associated with the entire life cycle of
biocomposites and conventional composites? How much of this energy is used
as feedstock and fuel energy? Is there a net reduction in energy usage for
biocomposites?

- What are the environmental emissions associated with the life cycle of
biocomposites and conventional composites?

- How do biocomposites compare with conventional composites in important
impact categories such as global warming, eutrophication, acidification, etc? If
biocomposites have higher impacts in some categories, what are the steps in

biocomposites life cycle which contribute the most to those categories?

57

4.2 Goal and Scope definition
4.2.1 Goal of the study

The ultimate goal of the present LCA study was to carry out a “cradle-to-
grave” analysis for Polyhydroxybutyrate (PHB)-Kenaf fiber and Polypropylene
(PP)-Glass fiber composites, which includes mass & energy flows related to
extraction of raw materials, intermediate products manufacturing, transportation,
use-phase and end-of-Iife or disposal phase. However, the actual analysis was
limited to “cradle-to-factory-gate” study, with the reasons to limit the study,

mentioned below in following sections.

4.2.2 Scope of the study
As specified in ISO 140405 standard, to clearly define the scope of the
study, the following items were considered:
- Functional unit for the products
- System boundaries of the products being analyzed
- Allocation method
- Data quality requirements

- Impact assessment and Interpretation methods used

4.2.2.1 Functional unit

To compare the LCA of different products or systems, a reference is
required so as to normalize the results. Therefore, such a functional unit ls
selected which is a measure of the performance of the output of the product or
system being studied. For composites, mechanical properties are the most

important performance characteristics and based on a feedback6 from a major

58

automaker, the modulus is the most important property considered while
selecting composites for structural applications. As discussed in the Introduction
chapter of this thesis, under the present NSF-Premise project, processing and
mechanical property evaluation of PHB-based composites was done and a
comparison of the tensile properties for PP (polypropylene)-Glass, PHB-Kenaf

composites is presented in Figure 4.3.

CI Tensile modulus
C . e Tensile strength

to
‘i
a:
T‘
a:
I

q

o

O
1
T

U
41

Tensile Modulus (GPa)

  

 

 

PP PP + 30% RAW PHB + 30% PHB + 5%PHB-g-
GLASS KENAF MA-r- 30%KENAF

 

O
t

Figure 4.3 Comparison of tensile properties for PHB - (30wt%)Kenaf and PP-
(30wt%)Glass composites

The tensile property evaluations above clearly indicate comparable tensile
modulus for PHB-(30wt%)Kenaf and PP-(30wt%)G|ass composites, and
therefore the tensile modulus was selected as the performance criterion to
compare these composites. Processing of these composites was done by
extruding polymer matrix and fibers to obtain composite strands, which were
pelletized and injection molded to obtain standard tensile coupons/dogbone
samples for mechanical property evaluations. Since reliable industry data

regarding processing of biocomposites to produce automotive components was

59

unavailable, data from composites processed in CMSC (Composite Materials
Structures Center) was used to estimate the processing energy requirements.
Therefore, tensile coupons (Figure 4.4) were selected as a functional unit, in

order to accurately describe the system boundaries for the current LCA study.

 

Thickness: 3.23 mm'

    

. . 4;.

Figure 4.4 Standard tensile coupon for PHB-Kenaf composite
The functional unit selected for the study was a multiple of tensile coupon;
and to get relevant normalized LCA data, the weight of the tensile coupons was

determined, and is presented below in Table 4.1:

Table 4.1 Average weight of tensile coupons

 

 

Component Average weight (grams)
PHB — (30wt%)Kenaf 10.8058 :1: 0.0093
PP - (30wt%)GIass 9.8050 :I: 0.0358

 

 

The slightly higher weight of the tensile coupons of PHB-Kenaf
composites as compared to PP-Glass composites was expected. Even though
Kenaf fiber has lower density (1.25 g/cc) as compared to Glass fiber (2.5 g/cc); it
is more than offset by higher density of PHB (1 .25 g/cc) as compared to PP (0.9
glcc). To accentuate the differences between the LCA of PHB-Kenaf and PP-

Glass composites, a multiple of 1000 was used for the functional unit. Therefore,

60

all the LCA results were normalized to the production of 1000 tensile coupons for
PHB-(30wt%)Kenaf composites (weight = 10.81 kg) and PP-(30wt%)Glass

composites (weight = 9.81 kg).

4.2.2.2 System boundaries

ISO 14041 standard7 clearly defines the need for system boundaries, so
as to determine which life cycle stages are included in the LCA, to which level of
detail the environmental flows are tracked, and to make sure that inputs and
outputs at the system boundary are elementary flowsi. A clear definition of
system boundaries helps in the comparison of different product systems without
any ambiguity.

Process flow charts along with the system boundaries for the products

being studied are described below in following subsections.

4.2.2.2.1 PP-Glass composite system boundaries

Polypropylene is one of the olefin polymers which is classified as
commodity polymer because large volumes of the polymer are produced
annually. It is obtained by free radical polymerization8 of the propylene monomer,
which is obtained by refining and subsequent cracking of crude oil. Therefore,
significant amount of crude oil is used as feedstock energy in the production of
polypropylene.

Glass fiber manufacturing9 consists of four major processing steps: batch

preparation, melting and refining, glass forming, and post forming operations.

 

‘ As defined in ISO 14040, “material or energy entering/leaving the system being studied,
which has been drawn/discarded from/into the environment without previous/subsequent
human transformation”

61

Details of these processing steps will be discussed in the LCI (life cycle
inventory) results part of this chapter.

The outputs from the production steps above, PP pellets and chopped
fiberglass strands are transported to the composites processing facility, where
composite parts are obtained by extrusion and injection molding. As an example
of use-phase, these composites are used as structural components in an
automobile. Since both the components of these composites are nondegradable,
at the end of use-phase these composites are either landfilled or incinerated and
the ash component after incineration is Iandfilled. The process flow chart with the

complete life cycle of PP-Glass composites is presented below in Figure 4.5.

4.2.2.2.2 PHB-Kenaf composite system boundaries

Polyhydroxybutyrate (PHB) is the most widely studied among the
polyhydroxyalkanoates (PHA) family of biopolymers and belongs to the polyester
class of polymers. PHB is currently being produced at pilot scale by a couple of
companies: 1) As per the latest Metabolix brochure”, “it has demonstrated
economic production of PHAs by fermentation of plant-derived sugars and oils,
and are in the process of making PHAs directly in crop plants.” 2) Biomer11
produces PHB by bacterial fermentation using sugar (sucrose) as a feedstock.

The current LCA study is based on the LCI study of PHB published by

Akiyama et al.12

, in which they have compared the fermentative production
processes for PHB using soybean oil and glucose as carbon source. The best
scenario highlighted in their publication is the use of soybean oil as carbon

sources and thus this scenario has been selected for the current analysis.

62

However, the soybean cultivation and soybean oil milling data was modified to
improve accuracy, and fertilizer use and run-off data was added to account for
important missing flows. The details of these improvements are mentioned in LCI
section of this chapter.

Kenaf is a warm season annual fiber crop which is successfully cultivated
in the southern states of the United States. For composites processing, long
fibers are preferred, thus only BAST fibers, which are obtained from the bark of
the kenaf stalk are used in the composites processing. Details of the kenaf
cultivation and processing are presented in LCI section of this chapter.

Transportation of polymer and fiber, composites processing, and use-
phase steps for PHB-Kenaf are same as those defined for PP-Glass composites.
However, end-of—life step for PHB-Kenaf is different from PP-Glass composites.
Since both the PHB polymer and the kenaf fibers are degradable, composting
was considered as the most practical way of disposing the PHB-Kenaf
composites. The process flowchart for the complete life cycle of PHB-Kenaf
composites is presented below in Figure 4.6.

For the current LCA study, the system boundary included processes from
extraction of raw materials till the processing of composites. The use phase for
the composites was excluded because of the current ambiguity regarding the
exact amount of composites used as structural composites in automobiles. The
end-of-Iife comparison for PHB-Kenaf and PP-Glass composites was not done
because of the data unavailability for use-phase and different viable disposal

processes involved for PHB-Kenaf (Figure 4.6) and PP-Glass (Figure 4.5)

63

compositesismi. However, biodegradability evaluation under controlled
composting conditions was done and the results are discussed in Chapter 5 of

the thesis.

4.2.2.3 Allocation method

Allocation methods are necessary in order to accurately distribute the
energy and material inputs and outputs in a process where multiple products are
produced. Such scenarios are common in any industry, thus the ISO 14041
standard requires a relevant allocation procedure to be used in the LCA study.
This standard recommends that allocation procedures should be avoided if
possible, by collecting additional information about the sub-processes involved
for individual products. If avoiding allocation is not possible, then partitioning of
flows can be done based on the physical relationships between co-products of
the system or economic value of these co-products. Additionally, this standard
recommends another useful approach for avoiding allocation, called system
expansion. In this approach, for a system with co-products allocation procedures
are avoided by assuming that product systems with an equivalent function have
the same environmental impactS[SA91. This approach has been effectively used by
Kim et al.13,14 for LCA studies of Ethanol production system and PHA production
from corn, where they have replaced co-products corn oil with soybean oil, and
corn gluten meal, corn gluten feed with a sum of corn cultivation and nitrogen in
urea.

For the current LCA study, mass based allocation was adopted because

of the data limitations due to which system expansion approach is not used,

64

complexity of the system involving polymer and fiber LCls, and use of mass

based allocation for the PP production LCI obtained from DEAM database.

4.2.2.4 Data quality

Time related coverage: to ensure collected data were not outdated, only
the most recent studies were referredrsmu.

Geographical coverage: wherever possible data pertinent to US
production was used and if relevant data were unavailable, modifications to the
available data were done so as to make them relevant to US industry.

Technology coverage: technology mix for the data provided has been
mentioned; instances where data was relevant to a specific site or derived from a

case study have been documented.

65

I[Crude oil Extraefioj [Mining of Raw Materials ]

 

 

and Refining (Sand, Limestone, etc)

 

 

 

naphtha

[ Batch mixing ]

[ Cracking J
[ Melting and Refining ] I
Prop;erie\, |

 

 

 

 

1
I
I
I
I

 

 

 

 

 

 

System
I Glass fiber Formirg I Boundary

Propylene |
polymerization [Size application and] l/

Drying
PP pellets

 

 

 

 

 

 

 

 

 

Textile Glass fibers

I

I

I

I

I

I

I

I

I

| [ Transfigt by ] Transport by I
l I I
I

I

I

I

I

I

I

I

 

 

 

 

Truck

 

 

 

 

 

 

 

I

[ Transport to

[ Extrusion ]

I Injection molding ]

disposal facility

I Landfill I‘fi llncinerator’

__________- [

 

 

 

 

 

 

 

 

 

 
 
  

PP-(30wt%)Glass
composite

 

 

 

 

 

Transport to ]

Structural automotive Landfill
component I

Figure 4.5 PP-Glass composites process flow chart and system boundary

 

 

66

 

  
  
 

 

        
  

 

   
 

Z" )Z
I f m I
I %
Soybean ; Kenaf ‘ I
I Cultivation Cultivation
%_
l |
'- HanaaingI I / 321:3;
I Soybean Oil //
PHA. I I /
fermentation I Processing and I/
Baling I
Centrifugation \
and Drying Kenaf BAST'fibers

    

 

PHB granules

r_J
‘ Transport by
Truck
Transport by

 

 

 

 

 

 

Transport to

composting facility

  
  

  

Injection molding

 

/I=H B-(30wt%)Kenaf -
composite

  

Composting

l
I
I
|
Extrusion I
I
|
l
I
|

 
 
  

 

I'___—_—______—_

____ _____J

 

I Structural automotive

component
I

Figure 4.6 PHB-Kenaf composites process flow chart and system boundary

67

4.2.2.5 Impact assessment and Interpretation methods used

For further understanding of the life cycle inventory (LCI) results,
emissions data are classified according to specific environmental impacts in
order to determine the significance of these impacts for LCI results5. International
Standards Organization (ISO) has laid out the requirements for reporting Life
cycle impact assessment (LCIA) results in ISO 14042 standard”. [SAM]

In the current study, midpoint impact methods were used because of their
comprehensive coverage of inventory flows as compared to endpoint categories
and more international consensus for midpoint models compared to endpoint
models”. To compute endpoint categories greater amount of detail is required
compared to midpoint categories, thus computing endpoint indicators might
exclude some inventory flows, which lack such detail.

Initially, for this study TRACI impact assessment methods", developed by

the US. Environmental Protection Agency and adopted in BEES model18

, were
used. These impact assessment methods are currently under revision, thus the
final impact assessment results were calculated using CML methods19 published
by Centre of Environmental Science (CML) at Leiden University. The impact

assessment methods selected are presented below in Table 4.2.

68

Table 4.2 Selected impact assessment categories

 

 

CML Impact Category Units

Acidification g 802 equivalent
Freshwater-Aquatic EcoToxicity g 1,4-dichlorobenzene equivalent
Depletion of abiotic resources kg antimony (Sb) equivalent

Depletion of the stratospheric ozone g CFC-11 equivalent

Eutrophication g PO43’ equivalent

Climate change (100 years) 9 CO2 equivalent

Human Toxicity g 1,4-dichlorobenzene equivalent
Photo-oxidant formation 9 ethylene equivalent

Terrestrial EcoToxicity g 1,4-dichlorobenzene equivalent

 

 

All the impact categories noted above are midpoint categories. Every
emission in the inventory analysis has associated effects, for eg. 802 emissions
from the system lead to smog formation, acid rain and thus causing respiratory
diseases and effecting crops, forests respectively. In this example, respiratory
diseases, effects on crops and forests are endpoint categories while photo-
oxidant formation and acidification are midpoint categories respectively. Thus,
midpoint categories are located somewhere in between the emissions (cause) -

endpoint (effects) link and this connection is described below in Figure 4.720.

69

motomoamo «5898 new ESEE 6:28.56 c6653 Eczooccoo 5V 23E

.4... i. wmm , . ..
_mEoEco:>:m

 

The selected impact assessment categories are described in greater detail
in impact assessment results section of this chapter.

As described in Figure 4.1, the interpretation phase follows the completion
of inventory analysis and interpretation phase of LCA. This phase allows the LCA
practitioner to analyze results, reach conclusions, explain limitations and make
recommendations, which can have a direct impact in areas such as product
development and improvement. The ISO 1404321 standard lays out the
requirements for life cycle interpretation, and this phase should include steps to
identify significant issues and evaluation of the results by steps, which may
include completeness check, sensitivity check, consistency check, etc.

For the current study, significant issues in the LCI and impact assessment
results were highlighted by contribution, dominance, and influence analysis. The
completeness of the inventory and impact assessment results was checked and
gaps in the data were documented. The sensitivity of impact assessment results
(computed using CML characterization factors) was determined by using different
nutrient-outflow allocation method for soybean cultivation and TRACI22 (beta

version 2.0) eutrophication potentials.

71

4.3 Life Cycle Inventory (LCI) Analysis

LCI is the phase of LCA, which is carried out once the goal and scope of
the project have been defined. However, this sequence of analysis is not
required to be rigid and ISO 14041 standard recommends that recursive process
should be followed in order to refine the system boundaries of the systems being
studied. This process was also adopted for the current study, first to modify the
functional unit from being mass-based (production of 1 kg of composites) to
volume-based (production of 1000 tensile coupons of composites), and second

to include runoff emissions from cultivation of soybeans for PHB production.

4.3.1 Data categories

The current inventory analysis was modeled using TEAMTM software.
Thus, the data category classification defined by this software was used for the
LCI analysis, and these data categories are classified as follows: a) inputs, which
include raw material, energy, and additional inputs; 5) outputs, which include
products, emissions to air, water, land, and recovered matter, waste categories;
0) remindersisms], which include energy indicators such as total primary energy
and intermediate flows (usually not visible) presented for greater understanding
of the LCI results.

The data categories as defined above use prefixes to differentiate flows,
with (r) representing non-extracted natural resources, (a) for air emissions, (w)
for water emissions, (3) for emissions to land, (ar) for radioactive air emissions,
(wr) for radioactive water emissions, and E for energy indicators. The energy

indicators included in the reminders data category are: Total Primary Energy,

72

Non-Renewable Energy, Renewable Energy, Feedstock Energy, and Fuel
Energy. These indicators are defined below.

Total Primary Energy: the sum of all energy derived from natural
resources, either burned as fuel (Fuel Energy) or used as raw material in a
product (Feedstock Energyismei);

Non-Renewable Energy: the fraction of total primary energy which is
obtained from non-renewable resources (eg. gasoline, diesel, natural gas, LPG);

Renewable Energy: the fraction of total primary energy obtained from
renewable resources (e.g. wood, wind energy, geothermal, solar energy);

Feedstock Energy: the fraction of the total primary energy which is used
as raw material in the products (eg. naphtha is used as raw material for
production of propylene, Figure 4.5);

Fuel Energy: the fraction of the total primary energy used as fuel for
various unit processes in the system (eg. gasoline & diesel used for
transportation, natural gas for production of steam).

As per the definitions above, the following relationship among energy
indicators was confirmed for the LCI results:

Total Primary Energy = Non-Renewable Energy + Renewable Energy =
Fuel Energy + Feedstock Energy

The results from the inventory analysis contained as many as 350 flows,
and not all of them contribute significantly to the selected impact assessment
categories. Therefore, significant flows were selected by contributiomsmq]

analysis of impact assessment results for PHB-Kenaf and PP-Glass composites

73

and the data for only these significant flows along with few additional fIOWSii[SAZI]
are presented as LCI results‘". The flows selected by contribution analysis are
presented below in Table 4.3. The percent contribution of individual flows to the
impact categories in the table below should not be compared across the
systems, for eg. (a) NH3 contribution of 47% to acidification impact for PHB-
Kenaf system cannot be compared with 0.03% contribution for PP-Glass system.
If the percent contribution from one of the LCI systems is missing for a flow, this

means that the specific flow did not have significant contribution to the impact

category, but included in the LCI results for comparison across the systems.

Table 4.3 Significant flows for PHB-Kenaf and PP-Glass composites (selected by
Contribution Analysis)

 

* PHB-KenafLCl Impact PP-Glass LCI Impact

 

 

 

 

 

 

 

 

F’OWS Contribution (%) Contribution (%)
CMLZOOO-Acldiflcatlon 7
(a) Ammonia (NH3) 47% 0.03%
(a) Nitrogen Oxides (NOx as N02) 18% 27%
(a) Sulphur Dioxide (802) 0.32%
(a) Sulphur Oxides (SOx as $02) 34% 73%
CMLZOOO-Aquatic Toxicity
(a) Arsenic (As) 1% 1%
(a) Beryllium (Be) 44% 50%
(a) Hydrogen Fluoride (HF) 9% 10%
(a) Nickel (Ni) 19% 21%
(a) Selenium (Se) 9% 10%
(a) Vanadium (V) 4%
(w) Barium (Ba++) 3%
(w) Phenol (CGHSOH) 4%
CMLZOOO-Depletion of abiotic resources
(r) Coal (in ground) 62% 23%
(r) Natural Gas (in ground) 32% 23%
(r) Oil (in ground) 6% 54%

 

 

 

“ These additional flows were selected based on their contribution as raw materials; and
emissions, which currently cannot be quantified by available impact assessment
methods, for eg. (w) Acids (H‘), (w) Water: chemically polluted, Waste (hazardous).

"' Impact assessment and interpretation phase calculations uses all inventory data;
however, because of large number of flows (350), contribution analysis is used to limit
the display of results to flows with significant contribution

74

 

Flows

PHB-Kenaf LCI Impact 7 PP-Glass LCI Impact

Contribution (%)

Contribution (%)

 

CMLZOOO-Depletion of the stratospheric ozone

 

 

 

(a) Halon 1301 (CF3Br) 13%

(a) Methyl Bromide (CH3Br) 72% 82%
(a) Methyl Chloride (CH3CI) 13% 15%
(a) Trichloroethane (1 , 1 ,1-CH3CCI3) 3% 3%

CMLZOOO-Eutrophication

(a) Ammonia (NH3) 30% 0.1%
(a) Nitrogen Oxides (NOx as N02) 13% 99%
(w) Nitrate (NO3-) 0% 0.1 %
(w) Nitrogenous Matter (unspecified, as N) 48% 0.2%

9%

 

(w) Phosphorus (P)

 

 

CMLZOOO-Greenhouse effect (direct, 100 years)

 

 

 

 

 

 

 

(a) Carbon Dioxide (C02, biomass) -83%
(a) Carbon Dioxide (C02, fossil) 183% 100%
(a) Methane (CH4) 0%
CMLZOOO-Human Toxicity
(a) Arsenic (As) 41% 36%
(a) Benzene (C6H6) 11% 20%
(a) Beryllium (Be) 3% 3%
(a) Dioxins (unspecified) 2% 2%
(a) Hydrogen Fluoride (HF) 29% 26%
(a) Nickel (Ni) 6% 5%
(a) Nitrogen Oxides (N0x as N02) 2% 4%
(a) Selenium (Se) 4% 4%
CMLZOOO-Photo-oxidant formation
(a) Benzene (C6H6) 4% 12%
(a) Carbon Monoxide (CO) 24% 50%
(a) Ethylene (C2H4) 50%
(a) Methane (CH4) 15% 30%
(a) Sulphur Dioxide ($02) 2%
CMLZOOO-Terrestrial Toxicity
(a) Arsenic (As) 39% 42%
(a) Beryllium (Be) 5% 5%
(a) Mercury (Hg) 44% 46%

(a) Nickel (Ni)

(a) Selenium (Se)
(a) Vanadium (V)
(s) Zinc (Zn)

4%
1%
2%
3%

4%
1%

 

Additional

emission flows

 

(w) Acids (H+)

(w) Suspended Matter (unspecified)
(w) Water: Chemically Polluted
Waste (hazardous)

Waste (total)

NA

 

 

NA

 

75

4.3.2 Glass fibers LCI

The glass fiber LCI analysis was conducted using data mainly from a 2002
study9 published by the Office of Industrial Technologies, US. Department of
Energy (DOE). This study presents energy and emissions data for the production
of various types of glass using current technology mix for production in United
States.

As mentioned in section 4.2.2.2.2, glass fiber manufacturing consists of
four main processing steps, and these processing steps until end of melting and
refining are the same for all types of glass (flat, container, blown, fiber glass).
Forming and post-forming processes for fiber glass vary significantly from other
glass types. A detailed fiber glass production process is presented in Figure 4.8
(reproduced from the Energy and environmental profile of the US. glass industry
document23,24,25).

The first step in glass fiber production is batch preparation, involving
mixing of raw materials for glass in the desired composition, which varies for
different types of glass. The composition for E-glass (also known as textile fibers,
glass fibers) is listed below in Table 4.4.

Table 4.4 Fiber Glass Composition (weight percent)9
Glass type Si02 Na2O CaO B203 Al203

 

 

Aluminoborosilicate:Iow-alkali(E-Glass) 54.5 0.5 22 8.5 14.5

 

 

 

 

 

 

76

 

 

 

 

Raw Material

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Computerized Weigting
and Mixing
Cullet ;
CW Computer Controlled
I ~ Charging
f I
Chemical Electrical
Boosting Boosting
L i i
i 3
Auto-
MOxygen Regenerafive Oxygen Recuperative Tapped
m A Funace W Fumaoe Electric
System System M alter
t It 6
Air F ;
(1W0) F . (1W0)
"‘9
f WP (154090)
0 i it
Rotary Steam or Air Mechanical Flame
' ' Blowing Draw'ng Blowing
I L I
i V
Binder Size
Application Application
Insulation 1 I Textile
Curing 500°F (2600) Drying
I I
ti
Automated _
Glass Product ‘— lnspectmglPadtagmg ._, Cullet

 

 

 

I I

I I

 

_J.

Batch
Preparation
and
Charging
Component

Melting!
Refining
Component

Forming
Component

Post-
Formlng
Component

Figure 4.8 Fiber glass production [Brown 1996, EPA 1995, EPRI 1988]

In general, glass batch contains formers, fluxes, stabilizers and optional

colorants. Formers are the basic ingredients for glass, with Silica (Si02), Feldspar

77

(as a source of alumina), and Boric acid (B203) used as formers for E-glass.
Glass products can be produced using only silica. Since the melting point of silica
is very high (1723°C), fluxes are added to lower the melting point of the batch.
For E-glass, soda ash (Na2C03 or Na2O) is used as alkali flux. To make glass
more chemically and physically stable, stabilizers are added, with limestone
(source of CaC03 I CaO) used as stabilizer for E-glass.

Electricity use is the main source of energy consumption for E-glass batch
preparation. Blending or batch mixing of raw materials account for the majority of
the electricity use in this step. The energy use for crushing, grinding, and sieving
of raw materials, before being shipped to the production site is not considered in
the DOE study. Particulate emissions are the major emissions from the batch
preparation process, and the following steps contribute to these emissions:
unloading and conveying, storage bins, mixing and weighing.

Melting and refining is the second step, after the addition of batch mix to
the melting furnace. This mix of raw materials is heated in the furnace to
temperatures ranging from 2600°F to 3100°F. Because of such extreme
temperature conditions, melting and refining is the most energy-intensive part of
E-glass production process. Such a high temperature is required to facilitate a
series of physical and chemical reactions including melting, dissolution,
volatilization, and redox reactions. Melting of raw materials is complete when
glass is free of crystalline materials.

The refining process (also referred as fining) follows the completion of

melting, and involves removal of gas bubbles from thebatch and molten glass,

78

homogenizing, and thermal conditioning of the glass melt. E-glass is produced
using recuperative furnaces and oxy-fuel fired furnaces, both furnaces use
natural gas as an energy source, the difference being the higher oxygen content
of air used in oxy-fuel furnaces, which improves the energy efficiency of these
furnaces since less amount of nitrogen is used in combustion (nitrogen absorbs
large amount of heat, and this heat is lost through nitrogen oxide emissions).
Additionally, oxy-fuel furnaces have significantly reduced nitrogen oxides
emissions when 100 percent oxygen is used for fuel combustion.

The majority of the E-glass production emissions are from the melting and
refining step, in the form of particulates, nitrogen oxides (NOX), sulfur oxides
(SOX), carbon monoxide (CO), and fluorides.

The third step in E-glass production is forming, in which molten glass is
converted into the final product. Continuous filaments of textile glass fibers are
formed by forcing molten glass through very small holes and collecting glass
filaments over a roller, and applying a coating of water-soluble sizing and/or
coupling agents. These coated fibers are then sent to post-forming and finishing
operations. The main energy requirement in the forming operations is electricity
consumption for running machine operations, fans, blowers, compressors,
conveyors, and additional equipment. Particulate matter emissions are the only
emissions reported for forming operations of textile glass fibers in the DOE study.

The final step for textile glass fibers involves post-forming and finishing
operations. Textile fibers coated with sizing and/or coupling agents are passed

through drying oven to remove moisture, and are subsequently sent to curing

79

oven to cure the coatings. Finally, the cured glass fibers are chopped to obtain
glass fibers used for making composites. The natural gas used in the ovens is
the only energy requirement for post-forming and finishing step. Combustion
products including particulate matter, NOx, CO from dryers and ovens are the
main emissions.

The total energy requirements and emissions (normalized to 1 kg of textile
glass fiber production), including the individual steps energy requirements, are

summarized below in Table 4.5 and 4.6 respectively.

Table 4.5 Total Energy requirements for 1 kg of Textile glass fiber production

 

 

 

 

 

 

Processing steps Electricity consumption (MJ Natural Gas
electricity/kg glass fiber) consumption (MJ/kg
glass fiber)
Batch preparation 1.34 NA
Melting and Refining NA 11.98
Glass forming 8.37 NA
Post forming and
NA 3.81
Finishing
Total 9.71 15.79

 

 

 

The data obtained from the DOE study only reported energy and
environmental numbers for the production step of glass fibers, and thus cannot
be considered as a complete LCI analysis. Based on the computed raw material

requirements (Table 4.4), energy requirements (Table 4.5) and emissions from

80

the production step (Table 4.6), the cradle to pellet LCI for production of textile
glass fibers was modeled using TEAMTM 4.0 LCA software”, with missing data

inserted from the associated DEAMT'VI LCI database.

Table 4.6 Emissions from 1 kg of Textile glass fiber production

 

 

 

 

 

 

 

 

 

 

 

Processing Emissions (kg/kg glass fiber)
steps Particulates SOx NOx CO Fluorides
Batch

1.90E-03
preparation ‘ NA >
Melting and

8.00E-03 1.50E-02 1.00E-02 5.00E-04 1.00E-03
Refining
Glass forming 5.00E-04 ‘ NA >
Post forming

6.00E-04 NA 1.30E-03 7.50E-04 NA
and Finishing
Total 1.10E-02 1.50E-02 1.13E-02 1.25E-03 1.00E-03

 

 

 

 

 

 

Modeling was carried out in two steps: first to develop LCI for mining of
raw materials and there transportation, and second for LCI of textile glass fiber
production. Mining of raw materials model was developed based on the raw
material requirements (Table 4.4) for 1 kg of raw material mix for glass fiber
production. This model is described below in Figure 4.9. Datasets for sand
quarrying, feldspar quarrying, limestone quarrying, and sodium carbonate

production were included from the DEAMTM database. Data for boric acid (B203)

 

“' TEAMTM acronym refers to Tools for Environmental Analysis and Management,
software developed by Ecobilan, PricewaterhouseCoopers.

81

production were unavailable in the DEAMTM database or other sources and[SA23]
hence not included in the inventory calculations. Transport by rail for 100 km was
assumed for all the raw materials shipped to the glass fiber production site and
associated energy use and emissions data obtained from DEAM database.

The Glass fiber production process was modeled based on the energy
and emission numbers computed for the DOE study and is described below in
Figure 4.10. DEAMTM datasets for Electricity, Natural Gas, Diesel production,
Natural Gas combustion, and Road transport by truck were included in the
modeling of textile-glass fiber production. Finished textile-glass fibers were
assumed to be transported 100 km in a 40-ton truck to composites processing
plant. Information regarding the source, year, and geographical coverage of the

datasets is mentioned below in Table 4.7.

82

5.: 3:99: 32 con: 86.0 .2 Boos. me 659”.

 

 

 

 

 

 

 

 

 

 

/_..__.. ...—z 2...... _: _::.:7.

 

 

83

 

s V‘

 

 

     

mmoooa couozooa ..on: 666.0 hoe Eco—2 ore 2:9“.

 

 

 

 

 

 

____:..T._.._.___ ...:_._i//T_,_.. , —

84

Table 4.8 shows the LCI data for the production of 1 kg of Glass fibers,

based on the model described in Figure 4.10.

Table 4.7 DEAM datasets used for modeling of Glass fiber-mining of raw
materials and production process

 

 

 

 

DEAM TM datasets Data characterization Sources

Sand Quarrying Energy production included (US Data) BUWAL Environmental
Series No. 132. 1991

Feldspar Quarrying Energy Production included (US Data) lbid.

Limestone Quarrying Energy production included (US Data) lbid.

 

Sodium Carbonate

One US site data for all inputs (source (1)).
One US site data for emissions and wastes

(source (2)). Energy Production (US data)

(1) OCI Wyoming
(2) T9 Soda Ash, Inc.

 

Electricity (US,

average) Production

Coal (56.6%), Natural Gas (8.6%), Heavy Fuel
Oil (2.2%), Nuclear (22%), Hydroelectricity

(10.6%)

EIA (1996, 95, 94), DOE
(1988, 83), DeLuchi
(1993), EPA (1996, 94,

85)

 

Natural Gas (US)

Production

Processes include recovery, glycol
dehydration, gas sweetening, pipeline transport

of natural gas.

EPA, AP-42 (1996). GRI
(1995, 96). EIA (1996).

DeLuchi (1987)

 

Natural Gas (US,

Industrial Boiler):

No waste or water effluents considered.

Heating value: 52 MJ/kg

EPA AP-42 (1998), EIA
(1994), GRI (1995),

 

Combustion Argonne National Labs
(1993)

Diesel (US) lncludes:- domestic and foreign crude oil EIA (1994), EPA (1993,

Production production. US refinery operations only. 85), DOE (1988, 83)

 

 

 

85

 

DEAM “1 datasets

Data characterization

Sources

 

Road Transport
(Truck 40 t, Diesel

Oil, kg.km)

Data from diesel truck fuel consumption and
emissions (max and average load: 40 000kg).
Actual load of truck: 24 000 kg & empty return

accounted. Diesel consumption with max load:

39 liter per 100 km.

ETH 1996. Anhang B:

Strassengutertransport

 

Rail Transport (Diesel

Oil, kg.km)

 

European Average data (1993-1994) for Diesel
oil consumption: 0.0056 liter/km.metric ton

(shipped). Diesel Oil Production: US data

 

Laboratorium fur
Energiesysteme ETH,
Zurich, 1996

 

86

AmmmEoE

 

 

 

 

o maven 0 Sun? 0 808; N38.“ News." a .80. 8.65 898 .n.
o Smog o 8.98... 8mm: ~9de 8me women.» a am. 52.38 .n.
o 52. mo-moe.m 8m. 5.» 89mm? $3 83 9...... a 6:8. 2895 E
o Sumac o ammo... 8mm? Remand some: some: a .2. 2:82 a.
o eomew 0 new»? 0 Sum: 308... 8.33 a 8:2. nEoEE< .n.
"839:0
o ammo a god o 80.0 68.0 can... an... 59.0 8588:: can;
0 ammo o Bod o 80.0 89o 8m... .9: :92. com: .995
o Sm. :3 o 8.83 0 808.4 come: meme: 9. .oo .3 5255 E
- . . - . - . . 38 c. .0 959m c.
o 8 m5 e o 808 e o 8 m2 N 8 mam m 808 n 9. ..onz. $8.5 sauce 3
. . .9596 e. .8082.
0 m8 o o o o o o 3.. e 9. 22860 52.8w 8
0 meme 0 Swan; 0 808.6 808.4. m3... 9. Genoa .... came 3
0 8mm: 0 Smog o Shag 8.53 New»? 9. Genoa .... __o E
o wood o «$23 o :2. 68.0 3n... 9. .2.an s. new .9262 E
. - . - . . . m .285
0 EN 0 o 8 m8 6 o 3 Mao N 6o 0 new a o. e. .808. 9.288: S
0 mere 0 o o o o 33 9. .66. 5829. E
0 wood 0 moms; o 80.0 Sec 2»... 9. Genoa .... .80 S
o 80.0 o o o o 0 mac... 9. some o.o< oeom
”8:9:
wwoooum Us: 3 “damask cozoaten. cozm—EEOU cases—wohm €032.09:- 2305
c2330.... 5:322 . . . case—52m 8E: 26:.
8...“. ease as. renew“... ..o 685 new .9582 new .2382 Scream 3...". case

 

 

 

 

 

 

 

 

 

 

26...”. 820 ..o 9. Co 8.8285 52.8 o. 28.9 as“. 8.20 .6 .3 we can»

87

 

 

 

team

 

 

 

 

 

 

 

o 8m. EN o 8.86.. o 2086 8.m8.N 888... o T18. Egon... .3.
o 8.84... o 2.8% o 582. 8.83 8.83 o E... 8.2 .3.
o o o o o o o o o .5. can a.
o 8.83 o 8.83 8.83 8.83 oo.meo.N 8.88 o 3. 5285.3 ......
. - . . . . 8.08:0
o 8.9: N o o. m8 o o 238 N 8.88 o 8.8. o o -5. 3 6556225..» ......
. . . . . . . .Now
3 8o o 8o o 8o o .8 o o8 o No. 5 4N... NN o on .69 8.8.86 .33.... a.
o o o o o o o o o. .88.. .8.on 58.8 a.
o 69mg... o 808.6 No.83 8.83 3.83 3.83 o .8. 8.8.8 a.
. . . . . . . . .N02 8
o : 6N. o 8o o .8 o .8 o 88 P :4 o 8. o. o .62. 826 888.2 E
o 8.8: o No.8; 8.8? 8.8m... 3.84... 30:... o ..z. .822 .n.
. . . . . o .698.
o 8.88 N o 8.8.. N o 808 o 498. N 3&8 N 8820 356.2 a.
. . . . . :85.
o 808 o o 38.. o o News N 8.84 o 8.8... o o 8.52m $56.2 ......
o vooo 8.83 88 zoo new... :oo «8... o .36. menses. .o.
o No.8: o 8088 8.86.. No.m..o.N 8.8.”... 88.8... o 81.8965. .9
o 308.6 o 8.83 o 3.83 8886 8.83 o $1.88.... 88.2.... .n.
o N783 o NVmeoN o NW8... N788 3-8.3 o .880. .8. con: ......
o o o o o o o o o .88. 82.8 a.
o :-moNo o 2-83 o :-m8.N 808.6 8.8.8 o 68.6883 8.65 .3
8s 88 o.o.o 8o.o Boo oooo 28 NFN o .oo. 8.882 888 a.
. . . . . . . m . 2.80.
o .8 oe o8 .. 8.. o 8.. 8.. mNm 9: oz. :8 o; 88 Noe. 8.65 898 .n.
a 3 8.5 .
eeeeoi ...! :50:
.285. cozoauen. coaeaano actuate... cease—30......
58:66:. .868: 8.832.. 8...: so...
.3... «use .5. connect .6 .683 one .282 one .282 3.285 .3... 86.6

 

88

 

85.9.

 

 

 

8. o o o o o o 8. 5.9. ...o .88.. 28.. €8.83
0 8. o o o o o 8. 8.9. .59.. ..3. €2.88.
00 0
E... o o o o o o E... ...... 3.0583
c 8... 9 N8 o 5.2 8.8 8.8 2.. ..98m. 88...". .98 m.
o S... o 3.8... o 8... 8.8 $8 ...... .85. 28388 m
o 8... o N8 o 8.8 2.8 8.8 ...... 85$ 2838mm 82 m.
o 8... 2... No... 2.2 8.» NSN 2.8. ...... ..98m .8“. m.
o 8... 2.9 :8 2.2- NE: SN 3N ...... .358 .888“. m.
o o 8.8...N o o o o 8.8: 9. ..o .085
- "805583...
c 8 F 2. o 83.8... o as... .8... 88 9. :52. 28>)
.. 8.8.8. o 8.88 o 383 8-88 8.83 9. .8888... 98....
F o o o o o o p 9. .2... 820
. . . . . 8.2.8
0 .8 o o 8o 0 o N8 o .8 o 8.. o 5.. 2.8.580 882. .3.
. . . . . .8588...
o 8.. o o 8.. o 0 N8 0 8 o .No a m 8.8.... 8888.0. .3.
o o o o o o o a m E. 8.2.808 .3.
o 8.88 o 3808 o 8&8... 8.8...N 8.88... m .1818. .88.”. ......
- . - . - . - . . . AZ 8 505038...
c 8. 8.. F o 2 8. m o : 8.. m 2 8. F 2 8.. n m .28... 8888.2 .3.
o 8.8; o 3.8....” o 8.8.... 8.88 8.88 a ..82. 92.2 .3.
.. 3. .25
882.. ...: :58.
.035. 5.852; :ozuanEoo 5:052“. 53052“.
53052.. .53“: 53052; 8...: 32...
.2... 320 35. .2825 ..o .286 26 .2352 one .232 3.2.32.8 .3... 820

 

 

23¢

 

 

 

 

 

 

 

 

89

4.3.3 Polypropylene LCI

The system boundary for polypropylene (PP) manufacturing has already
been defined under the goal and scope section of this chapter and was described
before in Figure 4.5. For the current study, PP manufacturing data was directly
obtained from DEAMTM database and is based on a cradle to pellet LCA of
polypropylene published by the Association of Plastic Manufacturers in Europe
(APME)26. PP pellets were assumed to be transported 100 km in a 40-ton truck
to composites processing plant. DEAMTM datasets for Road transport by truck
and Diesel production were used for the transportation step of PP-pellets, similar
to textile-glass fibers. The assumptions for these datasets are described before
in Table 4.7.

Since PP manufacturing data were directly obtained from DEAMTM
database, the modeling of PP transportation step was combined with the
modeling of PP-Glass composites, which is described in detail in following sub-

sections.

4.3.4 Kenaf fiber LCI
Obtaining processed natural fibers for composites processing, generally
involves 3 steps: 1) Cultivation and Harvesting 2) Transportation 3) Processing.
The material and energy requirements for cultivation and harvesting of
kenaf were calculated from a resource use and cost estimation study done by the
Dept. of Agricultural Economics at Mississippi State University in 199927 (see
Appendix 4.3.4 for details). The estimates provided in the study are for growing

kenaf in the state of Mississippi. Kenaf cultivation involves the following operating

90

steps: plowing (disk harrow & row conditioning), herbicide and fertilizer
application, seed planting, cultivation, and irrigation. Fully grown kenaf plants are
silage (forage) harvested, moved out of the field in a boil buggy, and finally
compressed in a module builder (module builder is a stationary piece of
equipment that compresses the harvested kenaf into a 32’x8’ module or block).
The compressed kenaf module is then transported to the fiber processing facility.
Transportation by 16-ton diesel truck for 50 km was assumed for shipping
harvested kenaf from farm to processing facility.

The performance rate and diesel fuel consumption rate for the tractor used
in each step are provided below in Table 4.9. The diesel fuel consumed was

calculated as a product of performance rate and fuel consumption rate.

Table 4.9 Diesel consumption rates for Kenaf Cultivation and Harvesting

 

 

 

 

 

 

 

 

Tractor Diesel Diesel consumed
_ Performance .
Operating steps srze consumption (gal/kg BAST
rate (hr/acre)
(hp) rate (gal/hr) fiber)v

Disk harrow (2 times) 160 0.078 7.72 8.63E-05
Row conditioning 160 0.053 7.72 2.93E-05
Seed planting 160 0.070 7.72 3.87E-05
Cultivation (early) 160 0.095 7.72 5.25E-05
Silage harvester 100 0.600 5.40 2.32E-04
Boll buggy 160 0.220 7.72 1.22E-04
Module builder 160 0.220 7.72 1.22E-04

 

 

 

 

 

 

" Converted to kg/kg BAST fiber value, using Diesel calorific value = 150.84 MJ/gal,
poiesel = 887.15 kg/m3

91

The Mississippi State University cost estimation study provided useful
information regarding the resources consumed by various farm operations.
However, information regarding total kenaf yield/acre, BAST fiber percentage of
the total yield, amount of fertilizers and herbicides used was not available in this
study and thus obtained from other publications. In a 1993 study, Webber28 has
presented kenaf yield and fertilizer use data for 5 kenaf cultivars. Out of these
cultivars. yield data for Everglades 41 was selected because of its greater
percent stalk production (78.5%) than leaf production and kenaf stalks having
mid-range bark to core ratio (0.52, bark content is 34.2%). Herbicide application
rates were calculated from a 1996 research report by Baldwin et al.29 and
TreflanTM (trifluralin as active ingredient) was selected as herbicide for kenaf
production. Kenaf fiber yield, fertilizer and herbicide application rate data are
presented below in Table 4.10.

Mass based allocation was used to distribute Kenaf fiber production data
between BAST and core fibers, with no allocation to leaf production. Such a
method of allocation is justified because of the economic value of the products
obtained. BAST fibers are directly obtained from the bark of the kenaf plant and
because of their long fiber length and superior mechanical properties compared
to core fibers; they are used as reinforcement in composites and paper
manufacture. Core fibers are obtained from the remaining portion of kenaf stalks
and are used as fillers in composites, animal bedding, oil-absorbent material for
spill cleanup30'3‘. The only use of kenaf leaves can be as livestock feed, thus

justifying allocation of production data between BAST and core fibers.

92

Table 4.10 Kenaf fiber yield, fertilizer and herbicide application rates

 

Allocating 34.2% to BAST

 

 

 

 

 

 

Yield IApplication rate kg/acre
fibers (kg/kg BAST fibers)
Kenaf Leaf yield 1092.66 NA
Kenaf Stalk yield 4775.31 NA
Ammonium Nitrate (NH4NO3, as N) 67.99 4.87E-03
Superphosphate (triple, as P205) 29.14 2.09E-03
Potassium Chloride (KCl, as K20) 56.25 4.03E-03
TreflanTM 0.61 4.37E-05

 

 

 

Tractor emissions from kenaf cultivation and harvesting operating steps

were calculated based on a 1999 Swedish study by Hansson et al32. This study

reports variation in CO, NOx, and HC (hydrocarbon) emissions for different farm

operations, because of different load requirements for these operations. 802 and

C02 emissions are assumed to be constant for all farming operations, with $02

emissions calculated based on the sulphur content in the diesel fuel. The

emission factors obtained from this study are presented below in Table 4.11.

These factors are measured from a standard 70 kW (94 hp) tractor used in

Sweden. For kenaf cultivation and harvesting (Table 4.9), farm operations were

performed using 160 or 100 hp tractor; the emission factors for these tractors

were assumed to be same as the 94 hp tractor used in the study.

93

Table 4.11 Emission factors for various farming operations (in g/MJ diesel)

 

 

 

 

 

 

 

 

 

Emission factor (g/MJ diesel)
Operating steps
CO; 802 CO NOx HC
Disk harrow (2 times) 74.6 0.094 0.042 0.897 0.016
Row conditioning 74.6 0.094 0.042 0.897 0.016
Seed planting 74.6 0.094 0.108 0.948 0.034
Cultivation (early) 74.6 0.094 0.076 0.747 0.030
Silage harvester 74.6 0.094 0.300 1.300 0.200
Boll buggy 74.6 0.094 0.300 1.300 0.200
Module builder 74.6 0.094 0.300 1.300 0.200

 

 

 

 

 

 

The emissions related to herbicide application were calculated based on
the emission factors specified in EPA AP-42 document33 for herbicide and
pesticide application. Detailed calculations are described below:

Formulation of TreflanTM (fromgawAgroSciences website“)
Active Ingredient: Trifiuralin = 43%

Inert Ingredient = 57%

Specific Gravity = 1.12

Vapor pressure of Trifluralin = 1.1x10“ mm Hg @ 20-25°C

Application method: preplant soil incorporated treatment

Based on this information, the uncontrolled emission factor for active

ingredient — Trifluralin is 52 kg/Mg (equivalent weight of active ingredients

volatilized / unit weight of active ingredients applied), and the emission factor of

94

VOC from inert ingredients is 38% of the total inert ingredients applied. VOC
emissions from active and inert ingredients were added; the calculated total
Unspecified HC air emissions allocated to BAST fibers were: 1.043E-02 g/kg
BAST fibers.

The LCI of Trifluralin was estimated based on the energy consumption
values per kg of agrochemicals produced (obtained from a 1998 NREL study“),

which are mentioned below in Table 4.12.

Table 4.12 Energy consumption values per kg of agrochemicals produced

 

 

Energy Inputs kg / kg of agrochemicals
Natural Gas 0.64
Coke 0.045
Light Fuel 1.89
Steam 16.5
Electricity (MJ eleC[SA25]) 24.4

 

Trifluralin represents 26% of the agrochemicals-mix, for which the energy
consumption values have been obtained from the NREL study. In the current
inVentory for trifluralin production, the same energy values (Table 4.12) were
assumed to represent production of trifiuralin alone. In absence of direct and
comprehensive data available for trifluralin production, this assumption was
useful in modeling trifiuralin production in TEAMTM software, as described in

Figure 4.11.

95

mmmooa 228—60.5 5.93:5. ..oh .802 3+ 059“.

 

 

 

 

 

 

 

U

 

 

 

 

 

 

 

 

1-2.9.. ..,...._....m:.__. ..l. h__.._.._._.. ._.__.r_._,:fl_._ .

 

 

 

96

The energy requirements for processing and baling of raw kenaf fiber
were obtained from Kengro Corporation”, a Mississippi based kenaf-processing
company. The average harvested kenaf fiber yield on there farms is 6 short-
tons/acre (5443 kg/acre). Before harvesting the crop is left on the field for three
months to facilitate dew retting, which aids in removal of unwanted bark material
from the kenaf fiber strands in the bark. Separation of harvested kenaf fiber is
done using equipment similar to stick machine, which is used for cleaning
unginned cotton, as discussed in a 1999 patent by Stover30. The average
separation capacity of this equipment is 9 short-tonslhr, with power requirement
being 170 kW/hr. The fiber separation process produces a maximum of 32%
BAST fibers, 62% core fibers, and the remaining is waste material. Maximum
separation is obtained under dry atmospheric conditions, with more waste
material produced under humid conditions. Mass-based allocation was used to
split fiber separation power requirements between BAST and core fibers.

Separated Kenaf-BAST fibers are baled using a horizontal baler, with a
baling capacity of 25 bales/hr (1 bale z 300 lbs) and power requirement of 30
kW/hr. Particulate emissions are the only significant emissions arising from the
processing operations for kenaf fibers, and are weather dependent. For the
current study, this emission factor was not calculated because of the unavailable
data.

Similar to Glass fiber LCI, Kenaf fiber cultivation, transport, and
processing steps were modeled using TEAMTM software, to obtain inventory with

“cradle to pellet” scope. The modeling was carried out in two steps. In the first

97

step, transportation of harvested-raw kenaf fibers for 50 km in a 16-ton truck was
modeled. In the second, inputs and emissions from cultivation, transport, and
processing steps were included (as specified in Table 4.9 through 4.11); the
model is described in Figure 4.12 below. Datasets for Electricity, Diesel,
Ammonium Nitrate, Potash, Superphosphate production and Transport by 40-ton
truck were imported from DEAMT'”I database for this model. The assumptions for
natural gas, electricity, diesel, and transport by 40-ton truck datasets have
already been specified in Table 4.7. Table 4.13 provides details for rest of the
datasets, associated with Trifluralin production, Raw kenaf transport, Kenaf
cultivation and processing models.

The LCI data for the production of 1 kg of Kenaf fibers is presented below

in Table 4.14, based on the model described in Figure 4.12.

98

©5389,... ucm toamcmc. .:o=m>_=:o Eon... 3:3. .2 .305. «EV 9:9“.

 

a

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

99

Table 4.13 DEAM datasets used for modeling of Kenaf fiber Cultivation,
Transport, Processing and Trifluralin production process

 

 

DEAM TM Data characterization Sources
datasets
Steam (100% Production of steam with 100% natural Ecobilan 2002

natural gas)

Production

gas, steam production model developed

by Ecobilan

 

Light Fuel on

(US) Production

Includes: - domestic and foreign crude oil

production. US refinery operations only.

EIA (1994), EPA (1993, 85), DOE

(1988,83)

 

Petroleum Coke

(US): Production

Includes: - domestic and foreign crude oil

production. US refinery operations only.

lbid.

 

Ammonium
Nitrate (NH4N03)

Production

Production by neutralizing nitric acid
(HNOa) with ammonia (NH3), unit
processes are: neutralization,
evaporation/concentration, prilling,
cooling/drying. Water and solid waste

emissions data not included.

The Fertilizer handbook for energy
data (multiplied by 85% to account
for process improvements) and
based on the stoichiometric
requirements; Air emissions data

from EPA AP-42

 

Potash (KCI)

Production of Potash (KCI as K20)

The Fertilizer Handbook, values

 

Production multiplied by 85% to account for

efficiencies over time
Superphosphate Production of Superphosphate (triple as J.L. Vignes, G. André, F .Kapala.
(Triple) P205) Union des Physiciens, Paris, 1994
Production

 

 

 

100

 

DEAM TM Data characterization Sources

datasets

 

Road Transport Data from diesel truck fuel consumption ETH 1996. Anhang 8:

(Truck 16 t, and emissions (max and average load: Strassengutertransport
Diesel Oil, 16 000 kg). Actual load of truck: 6000 kg
kg.km) & empty return accounted. Diesel

consumption with max load: 29 liter per

100 km.

 

 

 

10‘l

 

Table 4.14 LCI of Kenaf Fiber (cradle to pellet): Production of 1 kg of Kenaf Fibers

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Diesel Oil
Diesel Oil Production Raw Kenaf Electricity Kenaf
Kenaf Production Road (for Tractor Fiber Production Ammonium Cultivation
Flow Units “02:30" (used for 23:55]” use - Transport (used for Nitrate ”:32: n Sulp’erpjhos'phate Jrlguralln Transport,
(total) Road Truck 40 t) cultivation (Diesel fiber Production 0 r0 uc Ion ro uc Ion and
Transport) Truck 16 t) processing) Processing
harvesting)
Inputs:
(r) Coal (in ground) kg 7.27E-03 8.7OE-05 0 8.30E—05 1.45E-04 6.69E-03 1.62E—04 4.94E—07 1.14E-05 9.53E—05 O
(r) Iron (Fe, ore) kg 1.09E-05 0 0 O O 0 8.93E-06 5.37E—07 1.10E-06 3.1OE-O7 O
(r) Limestone (CaCOS, in ground) kg 5.69E-04 6.90E—06 0 6.58E-06 1.15E-05 5.28E-04 8.24E-06 2.25E—07 4.47E-07 7.63E-06 0
(r) Natural Gas (in ground) kg 9.22E-03 2.81E-04 0 2.68E-04 4.69E-04 4.76E-05 6.73E—03 1.13E—04 1.20E-03 1.09E-04 0
(r) Oil (in ground) kg 9.75E-03 2.56E-03 0 2.44E—03 4.26E—03 4.37E-05 2.54E—05 5.32E-05 2.81E—04 9.08E-05 0
(r) Phosphate Rock (in ground) kg 0.016 0 0 O O 0 0 0 0.016 O 0
(r) Potassium Chloride (KCI, as
K20, in ground) kg 4.11E-03 O 0 0 0 0 O 4.11E-03 0 0 0
(r) Uranium (U, ore) kg 3.12E-08 1.46E-09 0 1.40E~09 2.44E—09 9.66E-09 1.60E—08 1.04E-11 6.34E-11 1.94E-10 0
Water Used (total) liter 0.306 0.061 0 0.058 0.101 7.16E-04 0.076 2.58E-04 0.007 0.002 0
Outputs:
(a) Ammonia (NH3) 9 0.427 1.33E-05 0 1.27E-05 2.21E-05 4.84E—06 0.427 6.20E-09 1 .10E-08 2.11E-06 0
(a) Arsenic (As) 9 6.30E-06 9.06E—08 0 8.64E—08 1.51 E07 5.86E-06 2.22E—09 1.13E—08 1.90E-08 8.39E-08 O
(a) Benzene (C6H6) g 4.00E-04 8.81 E—05 2.40E-06 8.41 E-05 1.51E-04 2.18E—05 2.98E-05 2.21 E-06 3.92E-06 1.71 E-05 0
(a Beryllium (Be) 9 7.15E-07 9.08E-09 0 8.67E-09 1.51E-08 6.73E-07 1.22E—11 7.59E-13 1.53E-12 9.58E-09 0
a C ' ‘ _
gigmarbon D'Ox'de (002' g 1.76E-04 4.32E-07 0 4.12E-07 7205-07 1.72E-04 0 0 0 2.38E-06 0
ass)
(a) Carbon Dioxide (CO2, fossil) g 67.750 0.925 7.560 0.883 14.092 16.713 18.004 0.437 1.028 0.433 7.676
(a) Carbon Monoxide (CO) g 0.099 0.003 0.019 0.002 0.045 0.003 0.002 4.76E—04 4.13E-04 2.39E-04 0.023
(a) Dioxins (unspecified) 9 5.74E-11 7.03E-13 0 6.70E-13 1.17E-12 5.40E—11 3.33E-14 2.01 E15 4.13E-15 7.69E—13 0
(a) Ethylene (C2H4) g 1.10E-03 0 0 0 0 0 9.05E—04 5.49E-05 1.12E-04 3.16E—05 0
(a) Halon 1301 (CFBBr) g 1.96E-08 2.94E—1 2 0 2.81 E12 4.91 E-12 3.12E—14 2.01E-09 1.09E-08 6.61 E-09 7.04E—11 0
(a) Hydrogen Fluoride (HF) g 5.61 E-04 6.51 E-06 O 6.21 E-06 1.08E-05 5.00E—04 2.67E—07 2.66E-08 2.98E—05 7.12E-06 0
(a) Mercury (Hg) g 4.08E-07 5.62E-09 0 5.37E-09 9.37E-09 3.64E-07 1.46E-08 1.30E-09 2.43E-09 5.72E-09 0
(a) Methane (CH4) 9 0.169 0.005 2.88E—04 0.005 0.010 0.030 0.114 0.002 0.002 0.001 0
(8) Methyl Bromide (CHBBr) g 5.67E-07 6.94E-09 0 6.62E—09 1.16E-08 5.34E—07 0 0 0 7.59E-09 0
(8) Methyl Chloride (CHBCI) g 1 .88E-06 2.30E—08 0 2.19E-08 3.83E-08 1.77E-06 0 0 O 2.51 E08 0
(a) Nickel (Ni) 9 1.25E-05 8.19E-07 O 7.82E-07 1.37E—06 7.79E-06 7.56E-08 5.63E—07 9.45E-07 1.39E-07 O
(jiggtrogen Odees (NOx as g 0.480 0.004 0.098 0.003 0.162 0.053 0.032 0.002 0.004 0.001 0.120
(a) Selenium (Se) g 4.66E-06 6.07E—08 0 5.79E—08 1.01 E-07 4.84E-06 2.68E-08 5.52E-09 9.26E-09 6.27E-08 0
(a) Sulphur Dioxide (302) g 9.62E-03 0 0 0 0 0 0 0 0 0 9.62E-03
(a) Sulphur Oxides (SOx as 802) g 0.163 0.005 0.006 0.005 0.019 0.059 0.008 0.001 0.059 0.001 0
a ‘ _
( )Tr'Ch'Oroe‘hane ‘1 ’1 ’1 g 7.08E-08 8.68E-10 0 8.28E-10 1.45E—09 6.67E—08 0 0 0 9.48E-10 0

CH3CCl3)

102

 

 

 

 

 

 

Diesel Oil

 

 

 

Diesel Oil Production Raw Kenaf Electricity Kenaf
Kenaf . Road . . . . .
. Production (for Tractor Fiber Production Ammonium . . Cultivation,
Fl . Fiber Transport . Potash Superphosphate Trlquralin
ow Units . (used for . use - Transport (used for Nitrate . . . Transport
Production R d (Diesel l . . D' I f'b P d . Production Production Production d
(total) oa Truck 40 t) cu tlvation ( lese l er_ ro uction an _
Transport) Truck 16 t) processmg) Processmg
harvesting)
(a) Vanadium (V) g 1.08E-05 1.80E-06 0 1.71E-06 2.99E-06 2.43E-08 6.53E—08 1.56E—06 2.55E-06 6.60E—08 0
(s) Zinc (Zn) g 7.05E—07 O 0 O O 0 5.78E-07 3.50E—08 7.16E-08 2.02E-08 0
(w) Acids (H+) g 0.168 5.73E-10 0 5.47E-10 9.55E—10 1.14E—10 4.14E-06 9.40E-10 0.168 1.05E-10 0
(w) Barium (Ba++) g 2.42E-05 1.62E—09 0 1.55E-09 2.71 E-09 1.72E-11 2.81E-06 1.32E-05 8.04E-06 9.83E—08 0
(w) Nitrate (NO3-) g 8.20E-06 8.57E-07 O 3.40E-07 5.95E—07 3.48E—07 3.40E-07 5.55E-06 6.47E-07 2.92E—08 0
(w) Nitrogenous Matter _
(unspecified, as N) g 7.26E-06 9.14E-11 0 8.72E-11 1.52E-10 9.69E-13 3.82E-06 1.70E-06 1.71E 06 2.12E—08 0
(w) PhenOI (C6H50H) g 8.14E-05 2.09E-05 O 2.00E—05 3.49E-05 2.22E-07 3.65E—06 6.38E-07 4.30E-07 7.44E-07 0
(w) Phosphorus (P) g 4.75E-08 0.00E+00 0 0.00E+00 0.00E+00 0.00E+OO 1.07E-08 2.23E-08 1.41E-08 3.75E-10 O
Emsifififggecj Matter 9 2.405-02 4.94E—03 0 4.71 E—03 8.23E-03 5.25E—05 2.37E-03 1.16E-04 3.39E—03 2.39E-04 0
(w) Water: Chemically Polluted liter 2.17E-02 5.87E—03 0 5.60E-03 9.79E-03 6.37E—05 0.00E+00 7.25E—05 1.17E—04 2.07E—04 0
Processed-Kenaf BAST fiber kg 1 O O 0 0 0 0 O 0 0 1
Waste (hazardous) kg 1.31 E-OS 3.54E-06 0 3.38E-06 5.90E-06 3.76E-08 3.23E-09 5.19E-08 3.06E-08 1.24E-07 0
Waste (total) kg 0.021 4.66E-04 0 4.45E-04 7.77E—04 0.003 4.31E-05 0.004 0.012 5.86E-05 0
Reminders:
Triiluralin (C13H16F3N304) kg 4.37E-05 O 0 0 O 0 0 O 0 0 4.37E-05
Diesel Oil kg 8.69E-03 0 2.40E—03 0 4.00E—03 0 0 0 0 O 2.29E-03
E Feedstock Energy MJ 0.14 0.11 -0.10 0.10 0.01 0.02 1.78E-04 -8.86E—08 -7.07E—04 0.01 0
E Fuel Energy M} 0.96 0.02 0.10 0.02 0.20 0.23 0.32 0.01 0.06 0.01 0
E Non Renewable Energy M] 1.05 0.12 0.00 0.12 0.21 0.20 0.32 0.01 0.06 0.01 0
E Renewable Energy M] 0.05 1.18E—O4 0 1.12E—04 1.96E-04 0.05 1.42E—04 8.02E—07 2.02E-06 6.36E—04 0
E Total Primary Energy MJ 1.10 0.12 0 0.12 0.21 0.25 0.32 0.01 0.06 0.01 0
Electricity MJ elec 0.089 0 0 0 0 0 0.023 0.004 0.004 0.001 0.057
Raw-Kenaf BAST fiber kg 1 0 O 0 0 O 0 0 0 0 1
Transport: Road (diesel oil,
kg.km) kg.km 150 0 O 0 50 0 0 0 0 o 100

 

 

 

 

 

 

 

 

 

 

 

 

 

103

 

 

 

 

4.3.5 Polyhydroxybutyrate (PHB) LCI
LCA studies by Gerngross37, Gemgross and Slater”, Kurdikar et al.39,
have addressed the question of PHA’s (polyhydroxyalkanoates) being
sustainable; however these studies lack the detail, which is necessary in LCA
studies to reach a reasonable conclusion.

For the current study, the inventory analysis for PHB production was

based on the 2003 publication by Akiyama et al.12

, which provides a detailed life
cycle comparison study of polyhydroxyalkanoates produced from soybean oil and
glucose as well as petrochemical polymers such as LDPE, HDPE, PP, PS, and
PET. The LCI comparison by Akiyama et al.12 for poly(3-hydroxybutyrate—co-
5mol% 3-hydroxyhexanoate) [P(3HB-co-5mol% 3HHx)] (see Figure 4.13)
production using soybean oil as carbon source and poly(3-hydroxybutyrate)
[P(3HB)] production using glucose as carbon source is based on a simulation of

a full scale fermentative production using SuperPro Designer® simulation

softwareismei

933/” ‘i 9.?”7” ii
O/C\ /C C C
CH2 0/ \CH/

5

2
.9

(R)—3HB 0'05

(R)3HHx

Figure 4.13 Structure of Poly(3-hydroxybutyrate-co-5mol% 3-hydroxyhexanoate)
[P(3HB-co-5mol% 3HHx)]

104

Soybean oil is produced from soybeans in an oil mill, and glucose is
produced in a similar way by wet milling using corn as a feedstock.

As per Akiyama et at”, for P(3HB) production, theoretical highest yields
using glucose and Iinoleic acid (C18H3202, a main fatty acid component of
soybean oil) are estimated at 0.48 g-P(3HB)/g-C source and 1.38 g-P(3HB) lg-C
source respectively, as described in the stoichiometric equations below:

C6H1206 +1.5 02 —_" C4H602 + 2 C02 + 3 H20

Glucose PHB monomer

c13H3202 + 4.75 02 ———> 4.5 C4H602 + 2.5 H20

Linoleic acid PHB monomer

In their preliminary laboratory studies, they have obtained P(3HB) in the
cultured R. eutropha bacteria, with relatively high yields of 0.76 g-P(3HB)/g-
soybean-oil compared to experimental high yields of 0.3 to 0.4 g-P(3HB)/g-
glucose. Therefore, in their study they have simulated large-scale fermentation of
soybean oil in recombinant strain of R. eutropha bacteria culture to produce 5000
tonnes per year of P(3HB-co-5mol% 3HHX) copolymer, with yields varying from
0.7 to 0.8 g-P(3HB-co-5mol% 3HHx)/g-soybean-oil. The stoichiometric equation
used in the simulation assumes the biomass formula as C4,2H7,402No,7g, and is

described below:

105

0.81(C18H3102)3C3H5 + 29.06 02 + 0.79 NH3 ——’

Tri-linoleyl glycerol
(soybean oil)

C4_2H7.402N0_79 + 6.22 C4H602 + 0.33 cameo2 + 15.40 co2 + 17.11 H20

Cellular PHB monomer 3HHx monomer
biomass

The fermentation process, including the downstream unit processes to
obtain PHA granules is described below in Figure 4.14. As mentioned above,
fermentation of soybean oil is done by the recombinant R. eutropha and mineral
medium containing NH3, NazHPO4, KH2PO4, Na3PO4, MgClz, (NH4)ZSO4, and
other minor components required for cell growth. Cellular growth takes place
under nitrogen-limited environment till the nitrogen source is exhausted and is
followed by PHA production from soybean oil. The fermentation medium is kept
under constant conditions: aeration (aeration rate is 0.5 wm‘”), agitation (power
requirement of 1.0 kW/m3), compressor pressure of 2500 kPa (absolute), and a
ferrnenter temperature of 34°C. The fermentation process is exothermic in
nature, and a constant temperature is maintained by using cooling water. For
comparison, glucose as carbon source replaces soybean oil, while keeping the
same fermentation conditions. For both carbon sources, PHA granules are
assumed to be recovered at 95 wt% from cells, and fermentation performance
parameters (dry cell concentration, PHA content of cell-mass, PHA yield, and

fermentation time) are variable.

 

Vi Aeration rate wm = gas volume flow per unit of liquid volume per minute (volume per
volume per minute), where liquid volume = useful volume (or working volume).
Reference: BioDictionary, Bioengineering AG, www.bioengineering.ch

106

Once the fermentation is complete, PHA granules are extracted from the
cells by using SDS (sodium dodecylsulfate) surfactant pretreatment, so as to
dissociate cell membrane proteins from PHA. This step is followed by washing
with NaOCl (sodium hypochlorite), a strong oxidizing agent used to digest cellular
biomass. After each of these unit processes, PHA granules in the solution are
concentrated by removal of cellular biomass using disk-stack centrifuges and
finally dried by a spray dryer to obtain PHA granules.

In this simulation study, Akiyama et al.12 have considered eight cases for
PHA production using soybean oil and two cases using glucose as the carbon
source and estimated the cost for production of PHA. These cases are generated
by varying the fermentation performance variables. Both of the glucose cases
have higher life cycle energy usage than all but one soybean oil cases. Akiyama
et al.12 have identified two factors responsible for the higher energy usage in the
glucose—based PHAs than in soybean-based PHAs: higher energy required to
extract glucose, and higher amount of glucose necessary for the production of 1
kg PHA. Based on the cost estimate for PHA production by soybean oil, authors
have considered one of the median-cost case as the base case for LCI
comparison with PHA production from glucose and conventional polymers
including LDPE, HDPE, PP, PS, and PET. Performance variables for this
median-case along with selected glucose case are mentioned below in Table

4.15.

107

:ozmmEEgo .

 

 

0.x

 

«c0500.: 03.52000»:

9:35. 1
is: ioAla'v'tac.s

 

 

 

 

800m ..0 C0 0E0>E< 00:006.; 000020 530202: (In. 36 059“.

9.2025 «250.5 9.35
86:05 5:825:00 Oil. 0:35 20:5
0.8.3... as :21. 3..

 

 

 

.... 00.3530 <Im ,
\ b. l l ’ a I vi
. \ . o I c .a
0.23.6080 5280 ...H
11.0... . 53555.0". 5600558 ,
000.05 000
. 85% ._< ......l

lfiufizfifiee 0:83:50 \. me
me. :0 . _ x .5: 35:00
2. 5:325:00 .0 .m Ink

......fi
amt:

 

        

5302058 6:31
B: E I: I
, 602.1% hoes. .6532
.6006. 500182 ex:
86055 -.
:8: 05.2.0.0 / I .

to :3... a... 5.53.: I.

 

 

 

I

8:85 .2 ...: m . t ...
g s \. m

 

108

Table 4.15 Simulation conditions for Polyhydroxyalkanoates (PHA) production
using soybean oil and glucose as carbon source (reference: Akiyama et al. 2003)

 

 

 

 

Production conditions Operation conditions
Raw Cell PHA Raw PHA
. Product Time Yield Batches PHA Fermenter
material conc. conc. material (tons/yr)
(hr) (g/g) (run/yr) (tons/run) size (m3)
(g/l) (%) (tons/run)
P(3HB-co-
Soybean
5mol% 40 100 85 0.80 152 43.4 34.7 600 5095
C)"
3HHx)
Glucose P(3HB) 30 200 75 0.37 188 75.7 28.0 300 4978

 

 

 

 

 

 

 

 

 

 

 

Thus, the median-cost case or base case for PHA production from
soybean oil was selected for current LCA study. Akiyama et al., in their LCI study
have only reported stepwise/resource—wise total energy usage (MJ/kg-PHA), CO;
emissions (kg/kg-PHA), and unit values (MJ/kg-resource, kg COzlkg-resource).
This information was used to calculate the resource requirement per kg of PHA

produced, and an example of such a calculation is described below:

Electricity use (MJ)energy use = [%1§]x 3.6 % = 8.40 MJ-electricity
Electricity use (M3002 emissions = [%]x3.6 % = 8.38 MJ-electricity

[SA29]

The electricity use for fermentation step has an energy requirement of
21.99 MJ/kg-PHA, and 002 emissions of 1.28 kg/kg-PHA. Energy use and C02
emissions per kWh of electricity are reported as 9.42 MJ/kWh, 0.55 kg/kWh

respectively. Therefore, the electricity use (in MJ) for fermentation can be

109

calculated by dividing the energy requirement and CO2 emissions by their
respective unit values.

The calculations above clearly show a close agreement between the
electricity use value calculated from energy use and C02 emissions. Similar
calculations were done for every input use and, as a rule, the resource
consumption value computed from energy use was used to calculate raw
material use for the complete life cycle of PHA production, as described below in
Table 4.16. Akiyama et al.”, have not reported energy and emissions data for
some of the fermentation inputs including Na2HP04, KH2P04, Na3P04, MgCI2,
(NH4)2SO4, and other minor components because of unavailability of data.

Therefore, these inputs could not be included in the current LCI study.

Table 4.16 Raw material use per kg of PHB produced using Soybean 0i|

 

Raw material use per kg of PHB

 

Soybean 0il (kg) 1.296
Ammonia (kg) 0.024
Cooling Water (kg) 1475
Electricity (MJ-Electricity) 8.901
Process Water (kg) 7.895
Sodium Hypochlorite (kg) 0.491
Sodium Dodecylsulfate (kg) 0.222
Steam (kg) 3.024

 

110

4.3.5.1 Soybean Agriculture and Transport

Soybean cultivation data reported by Akiyama et al. are from a 1990
USDA-FCRS (Farm Costs and Returns Survey) study for soybean cultivation in
nine major states in US. And soybean oil milling data has been obtained from
1994 oil milling operations reported by ERS/USDA. These calculations are
shown below in Figure 4.15.

As shown in Figure 4.15, only energy use and CO2 emissions data have
been reported. For soybean agriculture there is significant ambiguity in the
publication regarding inclusion of energy use and C02 emissions from the
production of fertilizers and agrochemicals used in soybean cultivation. Also no
fertilizer run-off emissions have been calculated, where such emissions can have
significant eutrophication and human health impacts“. Excessive amount of N
and P nutrients in waterways promote over-production of plant biomass (algae)
and degradation of these biomass consumes oxygen, thus creating hypoxic
conditions (low oxygen concentrations). This process of over-fertilization of
waterways is also known as eutrophication, which effects local (lakes, rivers, etc)
and regional (lakes, rivers, and sea) waterways. Fertilizer run-off from corn and
soybean cultivation in US-Midwest is a major contributor to the increasing
hypoxic zone in the Gulf of Mexico, which is currently the size of State of
Massachusetts and considered ecologically dead during most summers, thus
effecting the aquatic ecosystem in this area. Excess amount of N and P nutrients
in drinking water can also affect human health by causing diseases such as Blue

baby syndrome and Non-Hodgkin's lymphoma“.

111

..0 .0 050.62. E 000030.00 0.00 9.5:: ..0 0:0 50020.30 :00:>ow 3.0 0.30....

 

05.00.50. 00 0:0 00: 35:0 .0.
53 0.000 £903 0.... :0 0002050 00? 00000.0 m

 

0:0 :0 E00 .000: w .00.... :00500 50002.0 0... ...
0.00.5050 ..00 .0. .n. .z. 00.5.0“. .000

 

3.0.08.0 .00.. .00“. .062. 52000 w
w

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

628.9935. 00.0 .8058. 2.0.58.0;
003020.955. mod 2050lL, 00:00:93.0: 00.0 ..0 It
800909.05. 00.0 .000 I: 60020-35... 5.0 05.800 I:
“Osgaamifiz mmd 02 .I.‘ «ostentax\0§ mNd 300 I...
020999525. No.0 :0 .03 .0:0_m0m.lv 80090.9..02 mm... .0005 I:
0.0:“. 8:09:93... 8... 02 It
Ea... f 0.8.35.3 305.2... f 08.995200
00 >905. 00 gm.
_. ..o 80.009. «0.0. H N00
:0 >006an as. H >905.
. ...... ..o .9 203.8 ...I .3 80.2.9. 3.0.
.300. ”0.0.... coon>om .000: .2 052:0...m< ...0 So... 8.0.5 N00
9. 8.. ...o 0000.60. , 9. 9.0 “0:08.60

 

 

 

 

$8.: 0.0....
9. omé ..00E>om

112

In case of soybean oil milling, 1994 data have been used. However more
recent and detailed data is available from other publications. Because of these
inconsistencies and lack of detail in data reported by Akiyama et al.”, soybean
agriculture and oil milling data were substituted from other publications, with
details mentioned below.

Inputs and emissions associated with soybean agriculture were obtained
from a 2005 NREL-Technical Report“, which primarily quantifies the
eutrophication potential for corn-soybean rotational production in Eastern Iowa
watershed region over a period of 13 years (1988-2000). This study focuses on
agricultural production in the State of Iowa, so as to be consistent with the
previous study regarding LCA of corn stover41 production in the State of Iowa.
However, a smaller subset region of Eastern lowa counties with an approximate
area of 50,000 km2 is selected because of its high corn and soybean productivity;
availability of agricultural production, resource consumption and erosion data;
and being part of three major watersheds, which have several years of water
quality data available. This data is utilized to develop and calibrate nutrient
leaching models for corn and soybean cultivation. The region selected in this

study is shown below in Figure 4.16.

113

88565-1 ...)me E0... 00000050.. 50020.30 :00n>om 0:0 E00 .2 :050: 030. 00.00.0m 0:0 059”.

        
   

.3 8....
«0.6255325?

£093

:02”. 25.00.02:

  

00:00.52. 3:300
:0 0309 .0.— :_ 00022.: 02<

300.53 00:20.03 20E_xo..&<

114

This study combines the analysis of fertilizer run-off emissions from corn
and soybean cultivation, because of the interdependence of run-off emissions
from corn and soybean cultivation. The study utilizes two different methods to
allocate nutrient flows among corn and soybean crops. The first allocation
approach is based on the study by Van Zeijts et al. (1999)42 addressing allocation
of fertilizer flows between various crops in the Netherlands. The assumptions for
allocating nutrient flows are as follows:

- Nitrogen (N) flows are allocated to the crop to which it is applied in that
particular year

- Phosphorus (P) and potassium (K) fertilizer flows are allocated over the
entire crop rotation, proportionate to the fraction of these nutrients in the
harvested crops

This allocation approach is relevant for P and K fertilizers, since the
excess amount of these fertilizers is absorbed in the soil and leads to slow
release of the nutrients over the entire crop rotation. In case of nitrogen
allocation, this approach is termed simplistic because of two reasons: 1) N
fixation by soybean crops, where fixed nitrogen becomes available during corn
cultivation next year thus providing excess nitrogen and, 2) the excess N applied
during corn cultivation is absorbed in soil and could become available during
soybean cultivation next year. Therefore, a second approach, which specifically
modifies the N fertilizer allocation is used.

The second allocation approach is based on the field data quantifying the

actual leaching rates from corn and soybean cultivation (Weed and Kanwar43,

115

1996; Bjorneberg et al.“, 1996; Bakhsh et al.“, 2002), and presents
approximately equal nitrate leaching rates for corn and soybean crops, despite
little or no N application during the soybean"ii cultivation.

Based on the details provided above for allocating nutrient flows, the
second allocation approach was selected for the current study, since it equally
distributes the nitrogen flows to corn and soybean cultivation; while first approach
exclusively allocates it for corn cultivation. The NREL study40 acknowledges the
large discrepancy between these two approaches, and assumes that the most
probable accurate allocation lies somewhere in between these two methods. The
study recommends that future research should focus on greater understanding of
nitrogen flows between corn-soybean system. Therefore, the variation in LCI and
impact assessment results by using first N-allocation approach was analyzed and
presented in the interpretation phase of the current LCA study.

The NREL study quantifies the nutrient flows over a period of 13 years
(1988-2000), which includes very wet (1993) and very dry (1988) years, and thus
allowed the inclusion of variability of crop yield and nutrient flows as a function of
rainfall in the nutrient flow model.

The nitrogen available in the agricultural system is transformed in various
ways. Ammonia fertilizer in the soil is converted to nitrates within a few weeks by
nitrification reaction, and taken up by plants, denitrified or leached into water.
Denitrification process converts nitrate into N2 and nitrous oxides (NO, N20) as
intermediate products under anaerobic conditions, and in the presence of organic

carbon. Products from denitrification process are released into atmosphere and

 

"" Soybeans are leguminous crops, i.e. there roots can capture/fix atmospheric nitrogen.

116

NOx or NH3 in the atmosphere is transformed quickly (within hours or days) by
ammonification process, and returns nitrogen to the soil as HNO3, NH4N03, or
(NH4)ZSO4. Additionally, some plants (such as soybean) fix atmospheric nitrogen
and convert it into organic-N, which is part of soil. Soil erosion removes some of
the fixed nitrogen and a fraction of this removed nitrogen is transferred to
waterways. The chemical reactions involved and the N-transformation cycle

(Figure 4.17) in the agricultural system are mentioned below.

Ammonification NHZHCONHZ + H20 ———* 2 NH3 + COZ

NH4+ +1.5 02 ——’ NOZ- + 2 H+ + H20
Nitrification
N02' + 0.5 02 —* NO3'

Denitrificatlon , 5 (CH2O) + 4 N03' + 4 H+ ———> 5 (:02 + 2 N2 + 7 H20

(intermediate steps involved are: NO3'—+N02'—>NO—>NZO—? N2)

117

8030-20-00.32 50.. 0800200.. 500.0 05.8.50 .020... 0 a. 0.98 0002:2000: 0009.2 n E. 050.“.

                       

#8:!!!-
..88. .15.. {8.0.1.8 380.8

00“

 

m
m
m

.§Sz.=oazo:0u .5 - -
.....s... 5...... ... .- ,

 

. . .4 l ... .2182. 32080... at 8
. . . .9122. 03:33.0; u
x 02.02.20 E

gee

....
9.950.020.

50.8300
oh

 

 

Phosphorus transformation in the agricultural system is less complex as
compared to Nitrogen cycle. P-fertilizers are sparingly soluble in water (1-2%)
and leaching of phosphorus is from this fraction in soluble form as
orthophosphate (PO43), mainly to surface waters. The majority of P-fertilizers
(70-95%) are absorbed by the soil and transferred to surface water bodies by soil
erosion. Phosphorus acts as a limiting nutrient for crop production or algae
growth. Hence, even a small P increase in the waterways can cause large
increases in the algae biomass production. The phosphorus transformation cycle
discussed in the NREL study ignores the contribution from atmospheric P,
because of its very small percentage in the atmosphere. This transformation
cycle for the agricultural system is presented below in Figure 4.18.

The NREL study has developed empirical models for individual N and P
flows as described in Figures 4.17 and 4.18. For the nitrogen cycle, NH3/NH4",
N03', N20 and NO)( (as NO) flows are included in the model for total nitrogen
(TN) flows. Phosphorus flows are modeled as total phosphorus (TP), with
fertilizer phosphorus (FP) as only inflow and outflows as P removed with
harvested crop and P leached to surface waters. Phosphorus absorbed in soil
and transferred by erosion to surface water bodies is not considered because
sorbed phosphorus is generally not accessible biologically. Potassium outflows to
waterways are not considered because the NREL study assumes that no

significant potassium inflows are distributed to air or water environments.

119

Somnméfidsmfii E0: 00030030.. E290 .0.:...:0.:m0 .0203 0 :. 0.0.8 5000:2200 029.0005. 0.20 059".

     

503.? 0:02.50

E30»... E02530

120

The empirical models developed are calibrated against the actual nutrient
flow data for the Eastern Iowa watershed region from the USGS National
Ambient Water Quality Assessment (NAWQA) and the NCAA/USGS Hypoxia in
the Gulf of Mexico study. TN flows to the Mississippi River are calibrated against
the actual flows from 1988-1998, and the calibrated model estimates cumulative
TN flows (from 1988-1998) within 1% of the actual cumulative TN flows. The
NREL study compares the calibrated model with actual flow data and estimates
from GREET‘“5 model (which assumes a fixed fraction, 24% of fertilizer-nitrogen
leaches to watenlvays), and calibrated model trends well with the actual flow data
with a few exceptions related to dry years of 1988 and 1989. This comparison is
presented below in Figure 4.19. Similar calibration of total phosphorus (T P)
model is done against the actual flows from 1988-1998 and calibrated TP model
has cumulative error (11 years) of +2.1%. The comparison of calibrated TP flows

against the actual phosphorus flow data is presented below in Figure 4.20.

121

8020-05-05002 0.0....
.02: 50.00.00... 0... 0. 030.. E .0. .08.: 50:0 0:0 0.00 .0300 5.; .000... E 00.05.00 ..0 08.00500 0:. 0.00.”.

 

 

 

 

 

 

 

bbcw .mma —. ..mmm —. .wmm _. 5ng .902 8mm _.
hr bl D P D c
.
V
ocean

'—

.Yriirllfi...‘ r si’i:1tr’iifl\\+rlé N
0086-. m

. Do
I/ \ / p
m

< oacom r W
S

N“.

\ < 4‘ GOOOON .Au

J

\ w

coccmu 11h

.0:::0E+zu=.\.vui .0. l <
want 0; :0 uwnanll’l

 

 

.n... 1 . a a £392..

122

.oomfim
-Smdsmmz So... 55. 33.3.5. 2. o. 26.. E a. 9% .38 5.; .26.: at 8.99.8 .0 comthEoo om... 9%:

cch mam? mag .39. Nam? camp «wow

u I I n u o

 

‘

r A \. ....“
I» N/ 7th saw
< M 7\ 83....

 

 

 

A

 

 

1%

/ \ 82:
g 894.. w
. N

.3

88. l

1 N V

.39: 2.58.. n: 335.301? 252 w

‘

Sat 3:26 855+ (

 

 

 

 

 

 

 

 

123

As mentioned before, this NREL study presents soybean production data
for 13 years (1988-2000) so as to account of weather variations, where such
variations can significantly impact the fertilizer nutrient outflows from the soil.
Therefore, in the current LCA study, 13 year-average data were used for
modeling soybean agriculture and transport in TEAMTM software. As specified in
the NREL study, soybean cultivation is part of corn-soybean rotation (90% of the
selected region has C-S rotation based on the actual data), with conventional till
and no corn stover collection. These assumptions are based on the actual
farming practices in Eastern Iowa watershed region. Data regarding energy and
emissions from the farming operations were obtained from Sheehan et al.
(1998)”. Farming input data reported by Sheehan et al. are from USDA’s 1990
Farm Costs and Returns Survey (FCRS), and relevant to farming in Iowa. It is
assumed that soybean plants uptake CO; from atmosphere while growing (by
photosynthesis); 002 is then sequestered as biological carbon in soybeans.
However, Sheehan et al., have not considered CO; sequestered in soybean
plant roots. In the current LCI study, sequestered carbon in soybeans was
accounted as C02 — biomass (see Table 4.18 below). For farming operations,
energy consumption is by use of diesel, gasoline, and motor oil for planting,
cultivation, and harvesting of soybeans. Diesel and gasoline consumption is
assumed to be for diesel and gasoline tractors respectively. Emission factors for
these tractors, as specified by Sheehan et al., are mentioned below in Table
4.17. Also herbicide use for soybeans is reported by Sheehan et al. and in the

current study it was assumed that all the herbicide use is in the form of Trifluralin.

124

LCI of Trifluralin production was modeled based on the data from Sheehan et al.
and details of the model are mentioned before in Section 4.3.4.

The 13 year-average data for soybean agriculture, which includes data
from farming operations, and fertilizer-nutrient flows is mentioned below in Table
4.18. As recommended by Sheehan et al., soybean yield data reported in the
NREL study was adjusted to account for the soybean seed use in growing
soybeans. Soybeans seeds require more energy compared to soybeans because
of additional storage, packaging and transport requirements. Therefore, soybean
yield was adjusted by subtracting 1.5 times the soybean seed use from total
soybean yield, so as to account for additional energy requirements of soybean

seedS[SA30].

Table 4.17 Emission factors for fuel combustion in Farming Tractors (obtained
from Sheehan et al. (1998), LCI of Biodiesel, NREL)

 

 

 

Emission factors (g/MJ)
Fuel type Hydrocarbons
CO NOx PM10 802 CH; N20 C02
(HC)
Diesel 0.085 0.32 0.89 0.041 0.12 0.0042 0.0019 75.5
Gasoline 0.2 1.14 0.63 0.0074 0.0046 0.032 0.0019 67.7

 

 

 

 

 

 

 

 

 

125

Table 4.18 (a) Soybean agriculture average data (1988-2000)-|nputs""l

 

Soybean agriculture, C-S rotation, conventional till, no corn stover collection

 

 

Inputs Units Average (1988-2000)
Land use ha (hectares) 1,319,608

FN (fertilizer nitrogen) mt-N 104,645

FP (fertilizer phosphorus) mt-P 14,598

FK (fertilizer potassium) mt-K 47,078
Trifluralin mt 7,070
Soybean seed use mt 78,526
Gasoline mt 26,057

Diesel mt 53,972

Motor Oil mt 2,253

 

 

 

Table 4.18 (b) Soybean agriculture average data (1988-2000)-Outputs"‘

 

Soybean agriculture, C-S rotation, conventional till, no corn stover collection

 

 

Outputs Units Average (1988-2000)
Soybean yield mt 3,857,989
Adjusted-Soybean yield mt 3,740,200

(a) C02 — biomass mt-COz —6,855,712

(a) NH3-N mt-N 26,398

(a) NO)( as N02 mt-N 1,946

(a) N20 as N mt-N 1,395

 

 

Vi“ mt = Metric Tons

 

"‘ mt = Metric Tons, prefix (a) signifies air emissions and (w): water emissions

126

 

Soybean agriculture, C-S rotation, conventional till, no corn stover collection

 

 

Outputs Units Average (1988-2000)
(a) HC (unspecified) mt 1,689

(w) TN — SW mt-N 43,929

(W) N03" - GW mt-N 3

(w) TP — SW mt-P 1,168

 

 

 

The next step after harvesting soybeans is to transport them to the
soybean crushing facility. Sheehan et al. have assumed a distance of 75 miles

for transporting harvested soybeans to the crushing facility. Additionally, it was

assumed that 16 ton-diesel trucks were used for this transportation step.

Based on the information provided above in Tables 4.17 and 4.18,
Soybean cultivation, harvesting, and transport steps were modeled in TEAMTM
software and are described below in Figure 4.21. For this model, motor oil
production data was substituted with lubricant inventory data, because motor oil

LCI data was unavailable. Fertilizer application for soybean agriculture included

ammonia (NH3), superphosphate (triple as P205), and potash (KCI as K20).

127

roam—.99 vcm .mcwmoZmI £232.50 cmogom .8 .322 ..Né 2:9”.

 

 

 

   

 

 

 

 

 

 

 

 

 

 

.7... __ ...... 2......)5—2 _ ... ...»...r.

 

128

 

4.3.5.2 Soybean Crushing

In the current LCA study, it was assumed that soybean oil is used as a
carbon source for fermentative production of PHB. Soybean oil is obtained by
milling/crushing of soybeans, as described before in Figure 4.15 and oil milling
data reported by Akiyama et al.”, was replaced with more recent and detailed
data reported by Sheehan et al.“, in a 1998 NREL study regarding LCI of
Biodiesel. Oil milling data reported by Sheehan et al., is based on a detailed
engineering study of a single soybean oil production facility in the southeastern
region of the United States". However, this data corresponds to production in
1981 and to improve its accuracy, authors have developed a detailed material
and energy balance for soybean crushing, which was critiqued by the industry
specialists (National Oil Processors Association-NOPA). Solvent extraction
process has been used by Sheehan et al., to obtain oil from soybean. This
process involves following steps: 1) Preparation of soybeans, which involves
removal of hulls, grinding and flaking of soybeans 2) Oil is extracted from grinded
and flaked beans using hexane as a solvent in a counter-current extractor 3)
Recovery of oil from hexane is done using multiple effect evaporators with
maximum solvent recovery and reuse, while the extracted beans are processed
by drying and grinding to produce saleable soy meal 4) The recovered oil is
finally washed with hot water, so as to separate gums from the oil and crude
degummed oil is shipped as final product. A simplified flow diagram of the

soybean crushing process is described below in Figure 4.22.

129

Soybeans Steam

Delivered H “"5
l Soy Meal

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Cleaned B P .
Natural Storage and S°Ybeans ean reparation Meal
Gas Cleaning (hull removal, L— Processing
grinding and flaking)
T Beans Meal
Steam V
\Hexane Oil
Degummed A Extraction /\
Soybean oil Hexane Steam Oil-Hexane
T l mix
__ Soybean Oil Wastewater Solvent Soybean Oil
Degumming Treatment A Recovery Recovery
Sttam wastewater
Crude Soybean oil /
Gums

 

 

 

\
/

Figure 4.22 Overview of Soybean Crushing Process

The material and energy balance developed by Sheehan et al., assumes
the soybean fiowrate of 94,697 kg/hr, which goes through initial cleaning to
remove dirt, trash, and moisture and final flow of 88,047 kglhr soybeans to the
extraction process. Since soybean oil and soy-meal are the only products from
this process, allocation of inputs, emissions and energy flows is done on a mass
basis of the products and output flows from the extraction process are shown
below in Table 4.19. However, amount of soybeans used to obtain 1 kg of oil was
calculated based on the soybeans delivered to the crushing facility not the
soybeans obtained after initial cleaning. Allocated raw material, energy inputs,
and significant emissions for soybean crushing are mentioned below in Table

4.20. Hexane is used as a solvent for oil extraction and is recycled by oil

130

recovery, solvent recovery and meal processing steps; where hexane air
emissions are due to vent losses from meal processing and solvent recovery
steps. Natural gas is used as fuel in the air-drying systems for removing moisture
from soybeans in the storage step. Emissions from burning of natural gas in the
air-drying system were calculated based on a DEAM TM dataset for combustion of
natural gas in an Industrial boiler. Steam is primarily required for bean
preparation and meal processing. Water emissions from the crushing process
were not considered in the current LCA study, since the majority of the water
outflow is due to steam condensates, with small contribution from soybean oil (~1
wt% of the total outflow), and therefore categorizing such flow as water

emissions in TEAM software was not possible.

Table 4.19 Co-products from the Soybean Crushing facility (from Sheehan et al.,
1998)

 

 

 

Products Hourly flow (kg/hr) Mass (%), for allocation purposes
Soybean Oil 16,072 18.25%

Soybean Meal 71,975 81.75%

Total 88,047 100%

 

 

 

Based on the input and emission data in Table 4.20, Soybean crushing
process was modeled in TEAM TM software and is described below in Figure 4.23.
DEAMTM datasets for electricity, natural gas, steam production and combustion
of natural gas in industrial boiler were included in the model for soybean crushing
process. LCI data for hexane production is not available in DEAMTM database or

from other sources, therefore not included in this model.

131

 

$82.". @5520 cmogom ..2 Boos. m3. 9:9“.

 

E

 

 

 

 

 

 

 

 

... .... _ ......l...l.._,__.._m.z_t. .....,..._;,.::J

 

132

Table 4.20 Allocated inputs and emissions to Soybean Crushing process

 

 

Inputs Amount/kg of soybean oil produced
Soybeans delivered (kg) 5.89
Allocated Soybeans (kg) 1.076
Hexane (kg) 0.002
Water (kg) 0.004

 

Energy Inputs

 

 

 

 

Electricity (MJ-electricity) 0.27
Natural Gas (MJ) 1.20
Steam (kg) 0.305
Emissions

(a) Hexane (grams) 1.85

 

Assumptions for majority of the datasets used for Soybean agriculture and
crushing process model have already been included in Table 4.7 and 4.13 of this
chapter, and Table 4.21 provides details for the remaining datasets, which

include gasoline, ammonia, and lubricant production.

133

Table 4.21 DEAM datasets used for modeling of Soybean Agriculture and

Crushing Process

 

DEAM TM datasets

Data characterization

Sources

 

Gasoline (US,

unleaded) Production

lncludes:- domestic and

foreign crude oil production.

US refinery operations only.

EIA (1994), EPA (1993. 90, 87, 85), DOE
(1988, 83)

 

Lubricant (US)

Production

Includes:- domestic and

foreign crude oil production.

US refinery operations only.

lbid.

 

Ammonia (us, NH3)

Production

 

Production of synthetic
anhydrous ammonia.
Natural Gas used as
feedstock (60%) and fuel
(40%), Natural gas
reforming process, no CO;

recovery.

 

Energy requirements data based on
Fertilizer Institute data for 1994-weighted
average for US plants. Air emissions: EPA
AP-42 process emissions and emissions
from a Natural Gas Industrial Boiler. Water
emissions: DEAM database, for NH3

production.

 

134

4.3.5.3 Polyhydroxybutyrate (PHB) LCI model

As discussed in the beginning of section 4.3.5.1, soybean agriculture and
oil milling data reported by Akiyama et al., was replaced with more recent and
detailed data and modeled in TEAMTM software, as described in previous
sections 4.3.5.1 and 4.3.5.2. Additionally, energy and 002 values for electricity,
ammonia, and steam usage were subtracted from the total energy and 002
numbers reported by Akiyama et al. (for PHB fermentation), because DEAMTM
datasets for these resources were included in the LCI model for PHB production.
LCI data reported by Akiyama et al., for production of these resources is relevant
to production in Japan, therefore replacement with datasets from DEAMW' makes
the current PHB LCI more accurate and relevant to production in US.

The final product from the PHB fermentation is in the form of granules and
it was assumed that PHB granules are transported 100 km in 40-ton diesel truck
to composites processing plant. Based on the raw material inputs (Table 4.16) for
PHB production and soybean oil milling model developed, PHB production
process was modeled in TEAMTM software and is described below in Figure 4.24.
Assumptions for DEAMTM datasets included in this model have already been
described before in Table 4.7, 4.13, and 4.21 of this chapter. Table 4.22 below
shows LCI data for the production of 1 kg of PHB, obtained from the modeling of

PHB production process in TEAMTM software (Figure 4.24).

135

 

3805 8.8355 min. .8 Boo—2 vmé 9:9“.

 

 

 

 

E

 

 

 

 

 

136

...mmo.

 

 

 

 

22 5.08 on... 8... :38 2.32 «4.3 2.39 a .80. 0.2.6.0 898 a.
. . . . . - .832.
0 2o 33. o 5mg 4 o 0:. o 803 a 8m a...“ m .30. 8.6.0 828 a.
o 238.» o 8.3a... 3mm: 80%... Sm. .3 3.93 a .8. E=.__.cmm a.
o 48... 808.... 805... wood ~85 82. a3... a 8:8. 88:8 a.
o 808... o 8&8... 803m 3.m8.o 808.. 308.» m .2. 2:82 ......
o 95.8 o 808.. new»: 3.3% «no... 83. a 8:2. «.855... a.
"833:0
8.: mm: o .8... 32. $3. Rod Sud. 8... ..92. 8...: 582.
«mm... o o o o o 0 «Nu... 9. mom
:9... o o o o o o 34... 9. .oOmz
o 8.33 o o o o o 8.33 9. .Ezoo. 29......
o 8mm? o 808.. 80mg 808.. 8.83 803.. 9. .20 .3. 52.55 c.
. - . m . 8:390 :. .09. mm
0 an. o o o o o o a! o v. 9.. 8.8.5 22898 c.
. . 3:305
o 9: o o o o o 0 S2 o 9. c.. 3.8”. 9289... c.
o Bod 0 So... 303.. mood 302.? «3... 9. €59... .... ..o c.
o 9:... 0 3m. 5a «Rd 8o... 82. 9.1a 9. 8.59.. .... 80 .9382 c.
. . . . . . €5.96
0 3.. o o 808 o 3.9m m 8o o 308 r 48 o 9. c. .808. 2985... ...
o KEEN o o 383 o o 8.9m... 9. .20 E. :9. c.
o 95¢ o 802... 33. 83 N8... «8... 9. 65o... .... .80 a.
"3:9...
8 ...o
8.832 536m c 3. 3.2:... :89.
air...” .0 9. _. .3: .0065. c2333.... 5:260... gazes—.2“. case—50.... c2323:— 8...: 26E
$.05... 9....an .335: . ..o .285 523.... 3.2.85 285:2 m1... .
.86an 536.... Ram

 

 

 

 

 

 

 

 

 

 

mzm .0 9. So 8.838... 52.3 o. 0.88. 3.1m 9.398918%“. .0. Nu... same

137

 

o 3.33 o 8.33 8388 8.33 5.3.... 3.33 ... .535. .88.. .3.
. . . . . . .2 3 63.83:...
0 m8 ... o ..-3. a 8.33 o 2-3.. . «.3... m .8... a. a .882 8888.2 .3.
o 8333 o .93....” 8.33... 8.33 8.3.... 8.3.... a ..82. 28.2 .3.
o 3.33 o 3938.. $3.... 833.. ..-33 3.3.... m .18. 5.3 .3.
o 8.3 o 2.3... o 8.33 338.. .33 a ...... 884. .3.
o 333.. o o 8.3.... o o 8.33.» m ...N. oc.~ .m.
o 8.33 o 8.33.. 83...... 8.33 8.33 3.3.... m .>. E.....8.8> .m.
. . . . . 3.085
o .3 ... e o 238 m o 8933 . 8.3. . 8.3.. a m -...... 8355.58. ......
o 3...... 83 83 8.... 33 ...3 83 m 8 x8. 88.0 55.....amwm
o 5.... o o o o o 5.... a wow. 8.55 5.8.5 ......
o 8.33 o 8.33 838.3 3.33... 8.3... 3.33.... ... .8. 5.5.3 .m.
o 83 83 3o... 83 «.08 o. .... 8.... m 8 .02. "8.5 cogmwu
o 8.33 o 8.3.3 8.3... 3.33 8.3.... 3.38... a ..2. 88.2 ......
o 33.... 0 83.0.3 0 3.33.. .938... 3.3.3 3 .525. 8.8.5 .38.... .m.
o 8.33...” o 833.... o 8.3.... 8.3.... 8.3.... a :35. 8.53 3.8.2 .8.
o m... 3.33 8.... 83 :3 .3... A ...... m .35. 888.2 .m.
o 8.33 o 8.38... 8.3.3 8.33 Sm. 3.. 8.38.4 3 31.8222 .8.
o 33 o 8.33 8.33 88... $33.. .8... m ...... 8.8:... :85... .m.
0 8.33.3 o «33.3 8.33 9-33 3.3... 5.33 a .35. .8. 581.8.
o .8... o o 8.... o o 3.... m .330. 88.2.8 ....
o 23....” o 3.3.... «.33... 8.3.... .73.... 83...... m .8588... 8.5.0 ......
o 88.. 3...... 8o... «.3 .8... 83 83 o .00. 8.55.2 525 .m.
...o
5.5.50... 3 3. 2:
......» Manda“... 88”.... . 5.852.. 5.852.. 5.852.. 5.852.. gonna”? .. 5
:85... 8.5.5 55...... ..o 88... 53.... 9.8888 «.555 Q... .. a. a ...
-833... .89...

 

:aogow

 

 

 

 

 

 

 

 

 

138

 

......9.

 

 

 

8. 38.... 8 8 .. 8 .. 383 ......9. ...o .88.... 8.... 888......
388 888 8 .. 8 8 . 8 8.8... 9. . 58.8
8 38.. 8 8 8 8 8 3.... ...... ...... 8 8.... 80 .282
8 888 8 8 8 8 8 88... 9. 80 .8382
00 o
.888 .88 8 .. 8.8 .. 8 88... ...... 858$
88 m .. .. 8 «.8 om. .. 83 8.. 8...... ...... 8.9m. .55.... ...... m
8 ...... 8 ....8 88 8.8 ....8 8.... ...... ..98m. 8.8558 ...
8... 88. o N... .3... 83 3.. 8.8 ...... 8.8m 8.83858 52 m.
8... ..8 ...... 88 .N... 8.3 88 ....8 ...... 8.8.. ...... m
8 88 8.8. ..8 .88. .....m 88 88 ..s. .985. 88.88.. m.
8 388 888 8 8 .. .. 88... 9. ..o .820
8 3.... 8 8 8 8 8 38.. 9. 888.8
83.. 8 .. .. 8 8 8 8a.. 9. ..o 88.8
.32....501
.. 8m... 8 3.38... 8 888 8.... 8...... 9. .38.. 8.8.5
.. 8.388 .. 8.33 .. 8.3m... 8.38.. 8.38... 9. .8838... 98>)
. 8 8 8 8 8 .. . 9. ...:3 e\.._..............m.3...
. . . . . 88......
8 3. 8 8 8.. 8 8 8.. .. 3-.. .8 N .... .. ...... 2.8.85 .283 .3.
. . . . . . .8588...
8 8m 8 .. 8.. 8 .8 .. 8.. 8 3.3.. . ...... .. 8 .88.... 88888 .3.
8 8.8 8 8 8.3.... .. .. 31.. ... .... 8.9.89... .3.
...o
55.5... .. 3 .8...

AXIIM hummus—WWW: 385. 6030300....— ccgagm cog—0:595 Cozuavnvsn— nonhuman”...— uzca .50.“—

.\._oem 8.....qu :88... ..o .82.. 88.8 3.2.8... «.55.... 3...

-883... 83.

 

:3ng

 

 

 

 

 

 

 

 

 

139

4.3.6 Composites Processing data

System boundaries for the current study were selected to include
composites processing steps, as discussed before in section 4.2 (goal and scope
definition) of this chapter, and LCA results were normalized to the production of
1000 composite tensile coupons (functional unit). Energy requirements for
extrusion and injection molding of composites were calculated based on the
actual processing data for these composites at CMSC (Composite Materials
Structures Center). The detailed calculations are described in Appendix 4.3.6

and summarized in Table 4.23 below.

Table 4.23 Composites processing energy requirement (MJ-Electricity) per
1000-tensile coupons

 

Amount of composites
Material EExtmsion (MJ) Em. (MJ)
processed (kg)

 

PHB-(30wt%)Kenaf 10.805 17.25 19.72

PP-(30wt%)Glass 9.805 13.55 15.93

 

Energy requirement for processing composites is only in the form of
electricity consumption for extrusion and injection molding steps. Higher
electricity consumption for PHB-Kenaf as compared to PP-Glass composites was
expected because of greater amount of PHB and kenaf fibers used for producing
tensile coupons and higher specific heat of PHB, kenaf fibers compared to PP,
glass fibers respectively; though this increase is partially offset by higher injection

molding temperature for PP-Glass composites.

140

4.3.7 Composites LCI models

Inventory analysis for production of PP-(30wt%)Glass and PHB-
(30wt%)Kenaf composites was conducted by combining the component models
described in previous subsections (Sections 4.3.2 — 4.3.5) and calculating energy
requirements for processing these composites (see Section 4.3.6). Similar to
component models, composite models were set-up in TEAMTM software and are
described below in Figures 4.25 and 4.26.

For PP-Glass composites model, PP production data was obtained from
DEAMT'“ database, referring to a 1997 study published by Association of plastic
manufacturers in Europe (APME)26. Similar to other component LCI models,
transportation step for PP-pellets was included in the current LCI model,
assuming a 100 km transport in a 40-ton diesel truck to the composites
processing plant. Electricity consumption data for extrusion and injection molding
of PP-Glass composites was included from the calculations mentioned in Table
4.23 above. LCI data for production of 1000 tensile coupons of PP-(30wt%)Glass
composites was obtained from the LCI model (Figure 4.25) and shown below in
Table 4.24.

In case of PHB-Kenaf composites model, PHB and Kenaf LCI data was
combined with electricity consumption for extrusion and injection molding data
obtained from Table 4.23 and Table 4.25 below, shows the LCI data for

production of 1000 tensile coupons of PHB-(30wt%)Kenaf composites.

141

3:89:00 8.2va

$39

in. so 88395 .e .585. m3. 2:3

 

r“. —

_v

 

 

 

 

 

 

 

d __...:I_.:_.__.:. #JIS: ...;

 

142

593858 amcoxAsEomEIn. .5 88:85 .2 .582 one 559“.

 

“Tl _ E

 

 

 

 

 

 

 

 

 

 

143

 

 

 

 

0 3.30.0 00mg; 00.m«m.m 3.m0«.m 0 .01. .0855. Am.

0 00«.0 «00.0 000.0 000.0 0 Ex. 00000.“. 00092.: .0.

0 :90: 0300.0 00.m«0.0 0300.0 0 0020. 009 8.01.0.

0 0 000.0 00: 03... 0 9:3. 0000.5 a.

0 00.m00.« 03.00.? 00.m«0.v 00.33.. 0 68.8.0000. 0006.0 .0.

0 003 .«00 000.3 000.3 0 60. 000600.... 00900 .0.

0 «0.002 0«0.0P« «.0080 0.03% 0 0.08.. «00. 0006.0 000.00 .0.

0 20.0 000.0 0.0«02- 0.0«000. 0 3000.0... «00. 0.00.0.0 00.000 .0.

0 3.30.0 00.m«0.« 00.95 3mm; 0 am. 05....000 .0.

0 000.0 000.0 :00 000.0 0 6:00. 008000 .0.

0 00-m00.« 00.m.00.« 8.90.0 03.3.0 0 .2. 0.000010.

0 «00.0 003 «00.5 0.0.«0 0 8:2. 0.00.002 .0.

"300000

0 0«0.0 «00.0 6:0 «003 .0... 00.0.. 08: 0000.5

0 0 0 «00.0 «00.. 9. wow

0 0 0 0:...” 0E.” 9. .0002

0 0 0 «0.0 50.0 9. .3100. 00.08:

0 033.0 «0.90.0 00.m.00.0 00.94.. 9. .90 .3 0.0.005 S

. . . €00.00

0 0 20.0 0«m 0 03 0 9. 0. .09. mm .60.. 0000.00 0.0.30.0“. S

0 0 000.0 003 03.. 9. 8000.0 0.. 0.8”. 0.0000000 3

0 0«0.0 «00.0 000.0 3.0.0 9. 6000.0 0.. ..o S

0 ««0.0 000.0 000.0 «3.0 9. 60090 0.. 00o .2302 S

0 0&0 «00.0 000.0 02.0 9. .00020 0. .808. 000.80.... S

0 0 0300.0 :00 30.0 9. .20 .0“: 02. S

0 03.0 «0.0 000.0 000.0 9. 8000.0 0.. .08 8

"8:9:

330.5 u: on. :o «co :

concaten— . ...—Wu. camp—Em. _ 5.82.2... 5302.2... :52. .04

8309030 .5: actuate... 03E 02.3. 3......" $.25. 3.32.30 3...: 30E
.008. $3. EoESE -2510... 000331..

 

 

 

 

 

 

 

000008 2.000. 0000 00 00.000020 ”02.00 0. 0.00.0. 9.000.000 .003. 0:0 ..0 .o._ 0«..V 0.000

147

 

 

 

148

0 0.0 00.0 00.«0 00.00 ...... 09000 000.0000“. 0

0 0 0«0.0 «0«.0 000.0 9. ..0 .0005

0 0 0 «00.0. «00.0. 9. 00000000

0 0 0 «00.0 000.0 9. ..0 0000000

”2000.50”.

0 000.. 000.0 000.0 0«0.0 9. 00.0.. 0.002.

0 00.000. . 00.00«.0 09000.0 00000.0 9. 000.0000... 0.002.

000.0. 0 0 0 000.0. 9. 3.80500 .008.-0:0

0 0«0.0 000.0 000.. 00... .0... 00.2.00 2.00.5000 00.0.5 .3.

0 0«0.0 000.0 000.0 000.0 0 80500000... .0002 000000000 .3.

0 0 00.000. 00«.0 00«.0 0 .0. 020000000 .3.

0 000.«0. 00.000.« 00.0.0.0 00.000... 0 $00.00. .00000 .3.

0 07000.0 00.000.« 000.0«. 000.0«. 0 .2 00 000.8000... 0000.2 000000902 0....

0 0908.. 00.000.« 000.0 000.0 0 .-002. 0.00.2 .3.

0 00.0«0.0 00.0000 00.0000 00.0«0.0 0 .100. 05000 .3.

0 00-00«.0 000.0 00«.0. «00.0. 0 ...... 00.00. .3.

0 0 00.00«.« 00-0000 00000.0 0 EN. 00.« .0.

0 0909.. 00.0000 00.0 .0.0 00.00.00 0 .>. .00.0000> .0.

0 00-0000 00-000.« 00.0000 00.00 ..0 0 0.00010... 0 . 0. 000500002000 .0.

0 000.0« 0«0.0 000.«0 0.0.00 0 .«00 00 .60. 00006 500.00 .0.

0 0 000.0 000.0 000.0 0 .«00. 0006.0 500.00 .0.

0 00.000. 00.0 .0. 00.0000 00.000... 0 .00. 05.00.00 .0.

0 .«0.0« 000.. 0«0.00 «00...: 0 .«02 00 x02. 00006 00000.2 .0.

0 000.0 00.0000 000.0 30.0 0 ..z. .0022 .0.

0 00.0000 00.0000 00.0.0. 00.000.« 0 00010. 0000.00 0.0.0.... .0.

0 00.000.« 00000.. 0.00000 00.0000 0 00.010.00.220 ......02 .0.

0 000.0. 000.0 00«.00 000.00 0 .010. 00050.2 .0.

00000.. a: 0:. :0 .00 0

00.8304... . 0....” 00.09....8— . coco-50...... 00:02.20 =an 5..

00000.....00 3.. 02.0002 0 .00.“. .008. 3.0.0 e0.0.00 0.. 0.00 3.0: 30....
.008. 0:0 0.00.8.0 60.020... .0080020

 

 

 

 

 

 

 

 

 

«00.000 0 .0..00« 0 000.000 00.9. .009. ...0 .0000. 0000 ”00000000

0 0 .0..00« 0 .0..00« 00.9. .020... ..00 000000..

08.0 o o o 30.0 9. .00.. 000.080.0200.

«00.« 0 0 0 «000 9. 000.0 000.0

00.0« 0 «00.0« 0 «00.00 00.0 ...2 0.00.8.0

0 00.0 00.00. 00.0.0 00.000 ...... 0.9000 0.0.000 .0000 0

0 00.0. 00.0. 00.0 00.00 ...... .9000 0.0030000 0

0 00.00 «000. 0«000 00.0.0 ...... 00.000 0.0030000 002 0

0 00.00 00.0. 00.00« 00.000 ...... .9000 .000 0

0 00.0 00.0 .0000 00.0«0 ...... 00.000 0.00.0000“. 0

0 0 000.0 0.00 0«0.0 9. ..o .0005
00320. 30.20:. 00.03.00. :82. .04

can-0:50.: Una c0335 eaten—99.n— comuoa—uOk— 3.009200 3.5... 32...—

00000500 .3: 00:02.2... 03.0. 30.0 32.32.50... 30.0. .

0006.00 3.00.8.0 00

 

 

 

 

 

 

 

146

 

E29. .__o .0923 081 £829.?

 

o o 238 3.88 3qu :29.

o o o 63% $93 9. 58%
Na.” 0 o o "a." 9. .8: 5% ficoxémmmoea
v8.5 0 o o 33. 9. 911m 3.03.0;sz

o o o 35.: «2.: 22 :3. mm 823 $0 Eamz

o o 0 R3 RN... 9. 80 .9352

33m 0 33 men: «no.8. 0% 22 @283

o Ea: 5m 8&8 8.8m 22 .656 55E .38 m

o 33 m; 53 8.3 :2 355 2855”. m

o 8.3 New, «3% 8.3... :2 >925 manicmm 52 m

o 8.3? m3 83v 3.3» 22 355 .8“. m

339:. 3520:. .333?-
:o=o:u21 new 5335 cozoauen. .thmyfihm hrnhwwpmwuw 8:5 26E
3.3an0 3.: cores—.2; 3...“. 3:3. 69¢an Eroxézn .
.23. 2.... 3.050%

 

 

 

 

 

 

 

149

4.4 l__ife chle Impact Assessment (L_C|£_Q

For a better understanding and comparison of the LCI results for PHB-
Kenaf and PP-Glass composites, impact assessment methods were used to
categorize inventory flows based on their environmental impacts, for eg. 002,
CH4, N20 air emissions are categorized under Greenhouse effect/Global
warming impact method, since they all contribute to global warming in varying
degrees. Impact assessment methods published by Center of Environmental
Science (CML)19 at Leiden University, were used in the current study, and are
listed before in Table 4.2. These methods are briefly reviewed in Section 4.4.1

below.

4.4.1 Impact Assessment Methods
4.4.1.1 Acidification
Acidification refers to the process of increase in hydrogen ion
concentration, [H‘], in terrestrial and aquatic ecosystems. This increase is
because of deposition (due to oxidation and hydrolysis) of air emissions, such as
NH3, N02, 802; and can cause reduced fish production in water bodies, effect
forests, man-made resources (buildings, automobiles), and human health.
Characterization method developed by CML is based on the RAINS
(Regional Acidification Information and Simulation) model and computes average
European characterization factor for acidification. Total acidification potential is
calculated by converting individual emissions into 9 SO; equivalents, as defined

below:

150

Acidification = ZAP,- x m,- .

I
where total acidification is expressed as 9 SO; equivalents, AP,- is the
acidification potential for air emissions of substance i, and m,- is the amount of
substance iemitted in grams.

CML acidification method and characterization factors are location
independent, for research study in Europe; but the current study focuses on
comparative LCA for US production. Still, this acidification method is used,
because it is based on an internationally accepted RAINS model and best

method available at the time of this analysis.

4.4.1.2 Human Toxicity, Freshwater-Aquatic and Terrestrial EcoToxicity

These impact categories refer to the impacts of toxic substances on
human health, aquatic and terrestrial ecosystems. The characterization factors
for these impact categories have been developed using USES-LCA (Huijbregts
1999“), which models exposure route, intake and fate of toxic substances
introduced in one of the environmental compartments (Global spatial scale has
three compartments: air, (sea)water, and soil; Regional and continental spatial
scale has six compartments: air, freshwater, seawater, natural soil, agricultural
soil, industrial soil).

Human toxicity potential for a single substance i emitted in compartment
ecomp is calculated by adding the contribution to each environmental
compartment via different exposure routes (for Human toxicity calculations, six

exposure routes are considered: air, fish, drinking water, crops, cattle meat, and

151

milk) and dividing it by a similar sum for a reference substance, 1,4-

dichlorobenzene. This calculation is described below:

ZZPDliecomsz X Ei,r X Ns

I' 8

Z Z PD'1,4-dichlorobenzene,air,r,s X E1,4-dich/orobenzene,r X Ns
r s

HTPiecomp =

where :

HTPLecomp = Human toxicity potential of substance i emitted to emission

compartment ecomp (dimensionless);
NS = population density at scale s;

PDli’ecomp, r. s 2 Predicted daily intake via exposure route r at scale s for

substance i emitted to emission compartment ecomp (day'1) and is a

combination of Fi,ecomp,fcomp , Tiifcomp’, and Ir;
Ei.r = effect factor, representing the human toxic impact of substance i, and

is calculated as reciprocal of the ADI (Acceptable Daily Intake) of the substance
via exposure route r (day);

Fj'ecomplfcomp = fate factor, representing intennedia transport of substance i

from emission compartment ecomp to final compartment fcomp and
degradation within compartment ecomp;

Tifcompl = transfer factor, fraction of substance i transferred from fcomp to

exposure route r;

I, = intake factor, representing human intake via exposure route r.

Based on the selected time span, total human toxicity is calculated as

follows:

human toxicityt = Z Z mi’ecomp xHTP;,ecomp,t
i ecomp

152

where :
human toxicityt = total human toxicity for time span t (kg);

HTP,-,ecomp,t = Human toxicity potential of substance i emitted to emission

compartment ecomp for time horizon t;

maecomp 2 emission of substance i to compartment ecomp (kg).

Freshwater-Aquatic and Terrestrial EcoToxicity impacts are calculated
similar to Human Toxicity, with difference being the use of PNEC (Predicted No
Effect Concentration) instead of ADI (Acceptable Daily Intake) as a reference
level for risk evaluation. PNEC is determined based on an extrapolation of
selected toxic effects for few species to entire ecosystem, so as to obtain
acceptable risk levels. Freshwater-Aquatic and Terrestrial EcoToxicity potentials

and total toxicity results are calculated as follows:

 

F AETP- _ PECi,ecomp,freshwater X Ei,freshwater
i,ecomp —
PEC1 ,4-DCB,freshwater, freshwater X E1,4-DCB,freshwater

TETP- __ PECi,ecomp,soil X Ei,soil
i,ecomp ‘-
PEC1,4-DCB,soi/,soil X E1,4-DCB,soiI

 

where :

FAETPi’ecomp = F reshwater—Aquatic EcoToxicity potential of substance i

emitted to emission compartment ecomp;

TETPLecomp = Terrestrial EcoToxicity potential of substance i emitted to

emission compartment ecomp;
PEC 2 Predicted concentration of substance i in freshwater or soil, due

to the emission in compartment ecomp, similarly PEC1,4-DCB predicts

concentration for 1.4-dichlorobenzene in either freshwater or soil;

153

E = effect factor representing the toxic impact of substance i or reference
material 1.4-dichlorobenzene in either freshwater or soil, calculated as

reciprocal of the PNEC (Predicted No Effect Concentration).

Freshwater-Aquatic EcoToxicity, = Z Z mi’ecomp x FAETP,-,ecomp,t

i ecomp

Terrestrial EcoToxicityt = Z Z mi’ecomp x TETPi,ecomp,t

i ecomp
where:
Freshwater -— Aquatic and Terrestrial EcoToxicity are total ecotoxicity results in

respective compartments for time span t (kg).

The CML study has selected infinite time span and global spatial scale for
its baseline Human Toxicity and EcoToxicity methods, and these methods were
selected for the current study because of their global scale selection, which
increases the relevance of impact assessment analysis for US production.

USES-LCA method presents a comprehensive modeling of fate of toxic
substances released in the environment, but has some uncertainties, which were
considered while interpreting the current LCA results and are mentioned below:

3) Weak modeling of fate of metals. Intermedia transport of metals is highly
dependent on environmental conditions and USES-LCA model does not
include any spatial difference for fate, exposure/intake, or effect parameters

b) Fate of geochemically reactive metals such as beryllium (Be) does not
include removal mechanism of hydrolysis (at salt water pH) and inclusion

into minerals such as ferro-mangenese nodules

154

c) USES model does not consider any variation in the safe level concentration
of some of the essential metals (e.g. Co, Cu, Cr, Mn, Se, Zn, these are
heavy metals), based on the original environmental concentrations of the
these metals

Therefore, Huijbregts48 (1999) recommends careful interpretation of the
results, in case of a significant contribution from heavy metals. This method
models the fate of 180 substances, which represents a small percentage of
known and unknown toxic substances and therefore it is possible that some of
the toxic substances in the current LCI are not included in the total toxicity

results.

4.4.1.3 Depletion of abiotic resources

Abiotic resources are defined as non-living natural resources, which
include for eg. iron ore, crude oil, natural gas, wind energy. Generally, abiotic
resources are classified into three categories: deposits, funds, and flows.
Deposits are resources, which are non-renewable during human lifetime,
examples are: fossil fuels, minerals, sediments, clays. Funds are resources,
which can be regenerated within human lifetime, examples are: groundwater,
soil. Flows are resources, which are continuously regenerated, examples are:
wind, river water, solar energy. Further distinction of abiotic resources is done by
classifying resources based on whether they can be depleted or competitively
used and thus deposits are classified in the first category, while flows are

grouped in the latter and funds may belong to either of these categories.

155

CML impact method for this category uses the method developed by
Guinee & Heijungs‘g, 1995, which calculates abiotic depletion potential based on
ultimate reserves and rates of extraction and the formula for this potential and
total abiotic depletion is shown below:

DRi x (Rref )2

ADP,- = 2 DR
(R,-) ref

 

Abiotic Depletion = ZADP, x m,
i

where:

Abiotic Depletion = total result in terms of kg of antimony;

ADP,- = Abiotic Depletion Potential of resource i;

m,- = amount of resource i extracted (kg);

R,- & Rref = ultimate reserves of resource i and reference resource,
Antimony (kg);

DR,- & DRmf = extraction rate of resource i and Rmf(kg/yr).

The method described above has few limitations: the depletion potentials
have been calculated only for elements and similar potentials for minerals & ores
are missing; this method only include resources which can be depleted, thus no
competitively used resources (all flows and some funds) are considered. Other

than these limitations, this method is time and location independent and thus it

was used for the current LCA study for US production.

4.4.1.4 Depletion of the stratospheric ozone
Stratospheric ozone depletion refers to the thinning (disappearance in
worst cases) of stratospheric ozone layer and is caused due to anthropogenic

emissions. This thinning of ozone layer leads to greater amount of solar UV-B

156

radiations reaching earth’s surface, which can have harmful impacts on human
health, natural resources, natural and man—made environment. In the CML study,
total ozone depletion impact and depletion potential for individual emissions are
based on a 1988 study by Wuebbles50 and the formula used to calculate these
results is described below:

5 [03 L’
5[O3]CFc-11

 

ODP,- =
Ozone depletion = ZODP, x m,-
i

where :

8[O3],. & 8[O3]CFC_11 = change in stratospheric ozone concentration at steady
state due to annual emissions of substance i and reference substance,

CFC-11 respectively;

ODP,- = ozone depletion potential for substance i,

in kg CFC -11 eq./kg of substance i;

m,- = mass of substance ireleased (kg);

Ozone depletion = total results expressed as kg CFC-11 equivalents.

The concept of ozone depletion potential (ODP) is similar to that of global
warming potential (GWP); however it differs in the way it calculates these
potentials, ODP’s are calculated at steady state, while GWP’s are calculated for
varying time spans (20, 100, 500 years).

The ODP’s used in the CML-Ozone depletion impact are based on the
best estimates of ODP’s published by World Meteorological Organization
(Wit/I0)51 in 1999 and 1992 and calculates ODP’s for an infinite time span

(approximated by steady state ODP). Additionally, CML-Ozone depletion impacts

157

are time and location independent, and therefore the current LCA study for US

production uses this impact assessment method.

4.4.1.5 Climate change

Climate change refers to the impact of anthropogenic emissions on
increased radiative forcing of the earth’s atmosphere. Radiative forcing is the net
difference between the amount of radiation coming into the atmosphere and
radiation going out; and net positive radiative forcing causes increase in earth’s
surface temperature, leading to changes in earth’s climate (also known as
greenhouse effect, global warming). Some of the critical endpoint impacts of
global warming include sea-level rise, increased droughts, forest damage, and
increased occurrence of extreme weather events, thus effecting all areas of
protection (human health, natural environment, man-made environment, and
natural resources). Example of greenhouse gases are: 002, CH4, and N20 and
global warming potential (GWP) of all the greenhouse gases is calculated by
normalizing it to CO; emissions, because C02 is usually the major component of
total global warming impact.

CML-Climate change method is based on a 1996 study by the
Intergovernmental Panel on Climate Change (IPCC)52, which has calculated
GWP for greenhouse gases with varying time spans of 20, 100, and 500 years.

The calculation for GWP and total climate change impact is shown below:

158

T
Ia,- -c,~(t)dt

GWPT', =7 0

[3002 ° CCOQ (t)dt
0

Climate change = ZGWPT’, x m,-
i

 

where:

GWPTJ- = Global warming potential of greenhouse gas i,
for a time span of T years;

Climate change = cumulative climate change impact,
expressed in kg C02 — equivalents;

a,- = radiative forcing per unit increase in concentration

of greenhouse gas i (w -m'2 -kg'1);

c,~(t) = concentration of greenhouse gas i at time t after the release (kg-m'

m,- = amount of greenhouse emissions (kg).

The CML baseline method calculates GWP’s for a time span of 100 years

and does not include net GWP’s for ozone-depleting gases (which have a

positive impact on global warming), because of the high uncertainties in these

net GWP’s. This CML-Climate change method is time and location independent;

therefore, it was used in the current LCA study to compute climate change

impact.

4.4.1.6 Eutrophication

Eutrophication refers to the impacts of excessive amount of nutrients (N &

P being the most important) in terrestrial and aquatic ecosystems. Over-

fertilization in aquatic ecosystems causes excessive biomass production (eg.

159

algae), thus creating hypoxic (low oxygen concentration) conditions and effecting
other aquatic species (fish, marine mammals, etc.). Additionally, high nutrient
concentrations in surface water may lead to surface water being unfit for human
consumption. Use of N & P fertilizers for crop cultivation is a major contributor to
eutrophication. The CML-Eutrophication method is based on a 1992 study by
Heijungs et al.”, which computes eutrophication impact for N & P compounds
normalized to PO43' eutrophication. The formula used to calculate eutrophication
potentials and total eutrophication impact is described below:

Vi lMi
Vref I Mref
Eutrophication = ZEP, x m,-

I

EP, =

where :

EPi -_—. Eutrophication potential of substance i;

Eutrophication = total eutrophication impact in kg PO43’ equivalents;

v,- & Vref = eutrophication contributions of one mole of substance i&
reference (PO43‘) respectively;

M,- & Me, = mass of substance i & reference (PO43‘) respectively (kg~mol'1);

The method developed by Heijungs et al.53, combines the eutrophication
potential for air and water emissions and does not consider any changes in the
overall result due to the sensitivity of the nutrient receiving environment and
limiting nutrient in that environment. Therefore, eutrophication results obtained by
this method are time and location independent.

However, their is significant variation in eutrophication results, if the

influence of location on atmospheric and hydrologic transport and deposition is

160

considered. TRACI‘7-eutrophication potentials developed by Norris“, considered
these influences, while calculating average eutrophication potentials for United
States. For water emissions, these potentials are similar to those calculated by

I 53.

Heijungs et a , while eutrophication potentials for air emissions have large

variation, as shown below in Table 4.26.

Table 4.26 Comparison of CML and TRACI Eutrophication potentials

 

Substance Heijungs et al. (1992), Norn's (2002), kg N

kg N equivalents/kg equivalents/kg

 

Air emissions

Ammonia (NH3) 0.83 0.12
Nitric oxide (NO) 0.48 0.07
Nitrogen dioxide (N02) 0.31 0.04
Nitrogen oxides (NO)( as N02) 0.31 0.04
Phosphorus (P) 7.29 1.12

Water emissions

Ammonium (NHH) 0.79 0.78
Nitrogen (N) 1.00 0.99
Nitrate (NOa’) 0.24 0.24
Phosphate (PO43) 2.38 2.38
Phosphorus (P) 7.29 7.29
Chemical oxygen demand 0.05 0.05
(COD)

 

161

Since significant variation exists between CML and TRACI air-
eutrophication potentials (Table 4.26), sensitivity analysis was done for CML
eutrophication impacts by comparing them with impacts calculated using TRACI

method (Section 4.5.2).

4.4.1.7 Photo-oxidant formation

This method covers the impact of formation of harmful chemical
substances (photo-oxidants) in troposphere due to the action of sunlight on air
emissions such as volatile organic compounds (VOCs) and carbon monoxide
(CO) in presence of nitrogen oxides (NOx). Some of the most important photo-
oxidants are ozone (03) and peroxyacetylnitrate (PAN), and exposure to these
compounds effect human health (e.g. asthma) and ecosystems (reduced plant
growth).

The CML photo-oxidant formation method is based on a 1992 study by

Heijungs et al.53

, which defines the total impact in terms of photochemical ozone
creation potentials (POCPs). The POCPs for this method, uses the
characterization factors published by Derwent et al.55 (1998), Jenkin & Hayman56
(1999), which combine the incremental ozone production due to incremental
VOC emissions with basic emission scenario (referred as ‘marginal’ approach)
and Derwent et al.57 (1996) (specifically for inorganic substances, including NO
and N02), which calculates POCPs based on the difference in ozone formation

with and without the particular substance (referred as ‘average’ approach). All of

these characterization factors use the data from a 5-day trajectory model of VOC

162

transportation above Europe, and assumes high background NOx concentration.
The formula used for calculating POCPs and total impact is mentioned below:

a; lb]
aC2H4 le2H4

 

POCP,- =

Photo — oxidant formation = ZPOCPI. x ml.

I

where:

POCP,- = photochemical ozone creation potential of substance i,

in kg ethylene equivalents/kg substance i;

Photo — oxidant formation = total impact expressed in kg ethylene equivalents;

a,- & 302H4 = ozone concentration change, due to change in the emission of

substance i & reference substance CZH4;

b,- & bC2H4 = integrated emission of substance i & reference substance C2H4;

m,- = mass of substancei released.

As described above, the CML method is based on a European VOC
emission model, with high background NOx concentrations and considered time
and location independent for a LCA study in Europe. This method was used for
the current study, because of the non-availability of a reliable impact assessment

method for US production.

4.4.1.8 Normalization and normalization factors

Normalization is an optional component of impact assessment as defined
by ISO 1404215 and is used to calculate relative significance of impact
assessment res'ults, check for inconsistencies, and prepare for additional

analysis, e.g.: grouping, weighting, life cycle interpretation. For the current study,

163

normalization was used to determine the relative significance of calculated
impacts. CML normalization method” and factors were used, so as to be
consistent with previous results. CML method calculates normalization factors
based on the yearly emissions of the substance in the reference region and
computing normalized impacts by dividing the total impact result with cumulative

normalization factor. The formula used for this calculation is mentioned below:

A658 = ZZZQexir X Mx,i.r.s
r i x

 

where :

A9,S = normalization factor for impact category e in reference situation 3

(kg equivalents yr'1);

Q“), = chracterization factor related to impact category 9 for substance x

emitted to compartment i in region r (kg equivalents kg'1);

Mx,i,r,s = annual emission of substance x to compartment i in region r

for reference situation s (kg yr'1);
S6 = total impact for category e (kg equivalents);
Ne,s = normalized impact for category e in reference situtation s.

This CML study reports normalization factors for four different reference
situations: The Netherlands in 1997/1998, Western Europe in 1995, World in
1990 and 1995 and reported normalized factors for all the impact categories
considered in the current study. This study recommends using global
normalization factors for regionally undefined LCA studies; therefore in the

current study global factors were used, even though the study was conducted for

164

US production of composites, using global factors is the best option rather than
using Netherlands or Western Europe factors, which can introduce a West-
European bias to normalized impacts. Thus, normalization factors for the World
in 1995 were used, since these are the most recent factors reported by the CML

study for global impacts. These factors are shown below in Table 4.27.

Table 4.27 Normalization impact factors for the World in 1995

 

Reference situation:

 

 

 

 

 

 

 

 

 

Impact Category Units
World, 1995
Depletion of abiotic kg antimony (Sb) equivalent 1.6 x 1011
resources yr"
Climate chan e 100
g ( 9 CO; equivalent yr" 4-1 " 1016
years)
De letion of the
p , g CFC-11 equivalent yr" 5-2 " 10"
stratospheric ozone
1 ,4-dichlorobenzene
Human Toxicity g . 1 5.7 x 1016
(DCB) equrvalent yr'
Freshwater-Aquatic g 1,4-dichlorobenzene 2 0 x 1015
EcoToxicity (DCB) equivalent yr"
1 ,4-dichlorobenzene
Terrestrial EcoToxicity g . 1 2.7 x 1014
(DCB) equrvalent yr‘
Photo-oxidant formation 9 ethylene equivalent yr" 9-5 " 1013
Acidification 9 SO; equivalent yr" 3-2 " 10”
Eutrophication g PO43' equivalent yr" 1-3 " 10”

 

Theoretically, the CML study has computed normalization factors using

region specific emission data and characterization factors. However, there are

165

impact categories, for which the emission data is extrapolated and/or
characterization factor for a different reference situation are used. Such
extrapolations may contribute toward the uncertainty in normalized impact factors
and the potential areas of concern have been well documented in this study. For
example:

- majority of toxic emissions to air, freshwater, and industrial soil for
Western Europe and World reference situations have been
extrapolated from emission data for Netherlands using GDP
extrapolation

- emissions of N and P to water and soil for Western Europe and the
World have been extrapolated from the data for the Netherlands,

based on human and animal population numbers

4.4.2 Impact Assessment Results

The impact assessment results were computed as a sum of product of
individual emission mass flows with their respective characterization factors for a
particular impact. As mentioned before in section 4.2.2.5, impact assessment
methods developed by Center of Environmental Science (CML)19 at Leiden
University were used to determine the impacts for current LCA study. For the
current study, contribution analysis of the impact assessment results was done to
determine the flows, which made significant contribution to selected impact
categories and only these significant flows are presented in the LCI and impact
assessment results, thus keeping the final results brief, but not missing any

important flows. This contribution analysis was previously presented in section

166

4.3.1, Table 4.3. Impact assessment results for PP-Glass and PHB-Kenaf
composites are presented below in Table 4.28 and 4.29 respectively. The results
include data about contribution from major production steps and contribution (%)

of individual emissions to the total impact.

167

 

 

 

 

 

 

 

$8... $8.3. $3.... $8.» 8.8.8 9. .88... .... ..o 3
$5... $8.8 $8.8 $8.8 8.83 9. 3:88 .... 80 .282 S
$8.8 $8.8 $8.8 $8.88 8.88.. 9. .888 .... .80 E
. . . . .00 000580. 050.00
.$8. .5 o .$: «8 0 .$8. 8 o 8 e .. >888 9. ..o 88.80.884.20 H 8.8....
$8... $8.8 $8.4 $8.” 8888 m .1880. .89... .3.
$88 $8... $8... $8." 8888 m .18. 588 .3.
$2.0 $3.8 $38 $8.» 8+8: .... S. 53.8.; 3
$8.: $8.... $8... $83 888.». m .8. 53.8.8 .9
$8. 8 $8... $8.3 $8.8 888... m ..2. 9.0.2 a.
$8.2 $F F...” $8... $8... 888... m E... 88...“. 88.2: a.
$8.8 $8.3 $8.8 $8.8 3+8: 8 .8. £2.88. a.
$8.. $8... $8.? $3... 588.... m .2. 2:82 a.
.$8. 8.2 .$ 5 8.0 .$ E 8.8 9.8 . moms“; m 2.0.on 883883.20 H 8.8.2
$8.8 $8.2 $3.8 $5.3 8+8? a wow 8 .60. 82x0 58.8 a.
$8... $83 $8.0 $8... 888.. m 800,. 8.6.0 58.8 ......
$8.8 $8. 8 $8.8 $8.2 , $80.... a .802 8 x02. 886 895.2 a.
$3... $8.8 $8.8 $8.5. 888.. m 812. 085.5. a.
.8: 8.8 .$: 8.” .$8. 888 8.2a .. .8 NOm a 8.88.2 .8835 H 8.8....
3.8.2.. 538.... 58:02.. :82. .0._
0:.a 00.0.5.6 0030300.... 3......» $.05... 8.000800 .80.“... :0: 3:5 30E
.5 :0. o: .— flca .
.Eoewoofl... 3.... e v. 690...»... 8:83.... 8:32:20

 

 

 

 

 

 

 

$2.00 0. 200.0. oumanoO .23. 9.5 .0. 83mm. Ememomm< 800E. owe 0.0m...

171

 

 

 

 

 

 

 

 

 

$88 $8.. $.90 $8.. 88.8.. m 80502.8... 8.5.0 .m.
$8..” $88 $3... $88 88.8.8 8 .8. 52.88 .8.
$8... $8.88 $8.. $8.8 88.8.. m .8200. 88:8 .m.
$8.8 $8.8 $ .8. $8.8 888.8 m .2. 8.82 .m.
. .$8. . . . .8 8.86.
0
...8. .o .8. 8.88 .$«.. 8 8. a. 8: . 000.... a 882888.20 H 88.2
$3... $8... o $8... .888 a . .80. 8322 .8.
$8.8 $8.8 $8.8. $88.88 888.. a ...»mo. .80. 8.8.0 880 .m.
. . . . .8050...
$8 o $8 0 o $8 a 88.8 . m .800. 8.8.0 .880 .m.
. .88”. . . .28.. 00.. 000020
00 . a t g N m .

.\ 8. 8 8.8 8 .88 .8... . 8.8. m. 8.8 00 m.mE__0-88._.20 . 850.2
. . . .z 8 .8888...
o o $8 .. $8. a 8.08 v m .882 8888.2 .3.
o o $8.... $8... 5.08.. m ..82. .28.“. .3.
. . . . . ~02
$8 8 $.. 8 $8 8 $8 8 3.08 . m 8 .62. 82.6 8822 ......
$8... $8... $8... $3... 6.08... a .812. 8.85.5. .m.
. . . . . 288.888
.8: 8 N .$8. 8 . .$.¢. B m 8 a. . .8 . on. a 88820 H 88.2
. . . . . 8.08:0
$8 8 $8 8 $8 8 $8 8 6.8. . a -.. .. .. 898226.; .8.
$8.3 $8.3 $8.3 $8.3 8.08.8 m ..0810. 8.830 ......ms. .8.
$8.8 $88 $8.... $8.8 5.08... m 0810.8.an 38.2 .8.
$8... $8... $8.. $8... 58.8.. m .880. .8. 8.8... .m.
. .$8. .$ ...... .8 808 88898.... m...

o . .
...8. 8.08 8 8.08... 898.8 3.0... . . ...0.0 m .o 58.80-880.20 . 8.82
$8... $8.. $8.8 $2.8 8.05.8 9. .88... .... ..o ...

32.22:
00.000? :88. .04
00.000090 00.000020 .300...— 00..
0:0 20.2508 8.000500 3...... 30.0
.0... 0 030:0 0.. 0 000.0 000.0 000.200.02.00 00... 0.. 0 0 «8.30220

0.8.8.0

 

 

 

 

 

 

 

169

 

 

 

 

 

 

$36 $86 $3.8 $86 «$83 6 E angm> E
$3.. $5.? $666 $3.. SE63 6 89 525.8 E
$-.v $65. $3.. $63 «6&6? 6 62V .222 E
$3.9. $33 $8.6 $2.8 3&86 6 a; 6:222 E
$5.6 $36 $86 $9.... 8&6? 6 39 £2.28 E
$66.? $3. 3 $~v6 $2.3 moms? 6 as o_=me< E
. . . . . .8 £2on
$69 3 6 $69 2 6 $2.. 8 3 6 3.2 . moo -3 6 5585366320 H 8522
$3.8 $866 $56 $66.3. 638.6 m Avxov 86522 E
$866 $66. 3 $866 $6 .66 «6.35 6 68 $685.2 8&8 E
$6: $3.: $26 $3.". smog 6 8:68 23:8 E
. . . . 6m :26th Emacs
$6: 2 6 $69 3 6 $5 2 6 k 6 . 225m. 6 5626666320 ” 8522
$E.v $666 $86 $36 3E6: 6 $9 525.8 E
$63 $36 $8.2 $36 8&6? 6 mm 63 89.6 ammonmwu
$2.6 $8.6 $26 $2.6 3&8...” 6 2 .322 E
$5.8 $8.2. $8.2 $8.3 8+3? 6 :5 $.82... 869%: E
652.“... . .
62.8.... .32 8..
case—52¢ coco-.3.“ .205“. no:
can 5.25.0 3.02.800 8:5 26....
.3: 52030.5 3...". 32¢ 2.230.529; anemia «5.30825

§ot§m

 

 

 

 

 

 

 

170

 

 

 

 

 

 

$8.6 $8.8 $8... $26 8+8: 6 EN. 2.8 E
$86 $8.2 $8.8 $8.. 888... 6 .>. 528:? E
$3.. $86 $86 $86 888.... 6 .8. 56.8.8 E
$88 $8.6 $8..” $8... 888.. 6 ..z. _e.o_z E
$8.8 $8.8 $8.8 $8.3 888.8 6 31:68.22 E
$28 $2.” $8... $86 888.. 6 .8. 52.28 E
$8.8 $3.8 $8.8 $8.8 888.2 6 as 0.826. E
Q E :6. 6.x 66. 26 s. 8. 8.8 . .8 .8.on
o o . 2 8 . 80 ..l 6 3.8888885 . 8522
$666 $86 $8.. $85 888... 6 Now. 8.85 52.6.8 E
$8.8 $3.6. $8.8 $8.2 8.868 6 :10. 9.2.22 E
$666 $8.8 $8.8 $86.. 8866.. 6 $80. 9.2.5 E
$8.8 $8.8 $8.8 $8.88 8.88 6 .8. 8.88... 598 E
$8.. $8.. $86 $8.... 23.8.8 6 6:8. 88:8 E
Q E 26 3 s 86 ... . . . .8 585.6.
a . $8 8 N 8 u . 82.5.... 6 2889268666820 . 8522
$8... $8.8 $66.” $8... 388.6 6 .8. 56.8.8 E
$8.. $8.6 $2.8 $8.. 6688.. 6 .802 8. x02. 886 863.2 E
$2.... $86 $8.... $8.» 3+8...” 6 ..z. .322 E
.853... 5.8....
2:. c2253 c2339.“. Ahhnmvfihm #fihww .32.“. co: 8...: 30.".
figmwwuwmi .3... .23. .88an 3.8.8.... 3.88820

 

 

 

 

 

 

 

173

4.4.2.1 Acidification

Acidification impacts were calculated in $02 equivalents and PHB-Kenaf
composites have 17% higher impacts compared to PP—Glass composites for the
same functional unit of production of 1000 tensile coupons and the higher
impacts are due to nutrient runoff from soybean cultivation and higher electricity
consumption for PHB-Kenaf composites. Comparison of these composites and

stepwise contribution towards acidification impact is shown below in Figure 4.27.

 

CML - Acidification
35° ‘ ' #um

IF'ber Prediction
| lElectricity—Conposies
Pmcessng
200 4

PP-Glou Oonpooln FIB-Kano! M

g 802 «1.

 

 

 

Figure 4.27 Acidification impact for PP—Glass and PHB-Kenaf Composites

174

The processes, which contribute the most towards this impact were
determined using contribution analysis. For PP-Glass composites, these
processes are:

- PP production = 47%

- Electricity production (for all sub—steps) = 26%

- Glass fiber production process = 23%

- Natural Gas production & combustion (glass fiber production) = 4%

In glass fiber production process (Table 4.6), major contribution is due to
SO)( and NOx emissions from melting and refining step, with small contribution
due to NOx emissions from post forming and finishing step. Additionally, there is
a small contribution from natural gas use and its combustion in glass fiber
production in the form of NOx emissions. Overall, for the complete life cycle of
PP-Glass composites, acidification impacts are due to 80x and NOx emissions.

Based on a similar contribution analysis for PHB-Kenaf composites,
following processes have significant contribution towards acidification:

- Soybean cultivation (nutrient runoff) = 50%

- Electricity production = 42%

- Superphosphate production = 2%

- Steam production = 1%

In case of soybean cultivation, NH3 and NOx emissions (Table 4.18) cause
acidification. As in the case of PP-Glass composites, electricity production
(composites processing and for PHB production) has significant contribution to

the total acidification impact because of NOx and SO)( emissions.

175

4.4.2.2 F reshwater-Aquatic EcoToxicity

Freshwater aquatic ecotoxicity (FAET) impact in the current study was
computed in 1, 4-dichlorobenzene (DCB) equivalents. Impacts for PHB-Kenaf
composites were 115% higher compared to PP-Glass composites, primarily
because of electricity consumption for PHB-Kenaf composites, which was
roughly twice the electricity used for PP-Glass composites. Cumulative impacts
for PP-Glass and PHB-Kenaf composites, highlighting stepwise contribution are

shown below in Figure 4.28.

 

CML - Freshwaer Aquatic EcoToxicity

     
 
 

4o 6

35 4 ~ -
IPotymerProduzuon

30 4 .
IFiberProdudIon

a 1.4-ace oq.
8

7 IElectricity-Cormosites
Processing

 

PP-Gllu Comm FEB-Kenaf Composi-

 

 

 

Figure 4.28 Freshwater Aquatic EcoToxicity impact for PP-Glass and PHB-Kenaf
Composites

For PP-Glass composites, the air emissions of Be, Ni, Se, HF were most
significant and following processes contribute towards the majority of FAET

impact:

176

- Electricity production (composites processing and glass fiber
production) = 96.5%
- Natural Gas production & combustion (glass fiber production) = 1.5%
In case of PHB-Kenaf composites, air emissions of Be, Ni, Se, HF are
significant, along with small contributions due to Vanadium air emissions and
water emissions of barium and phenol. The following processes are main
contributors towards the FAET impact for PH B-Kenaf composites:
- Electricity production (composites processing, PHB production
process, soybean crushing and cultivation) = 85%
- Steam production (used in PHB production process and soybean
crushing to obtain soybean oil) = 6%
- Miscellaneous steps associated with soybean and kenaf cultivation

(tractor use, transport by truck, fertilizer production) = 10%

4.4.2.3 Terrestrial EcoToxicity

Terrestrial EcoToxicity results measured in 1, 4-dichlorobenzene (DCB)
equivalents were similar to the results obtained for Freshwater-Aquatic
EcoToxicity, in terms of comparative impacts for both systems and processes
with major contribution to the total impacts. For both the systems, air emissions
of mercury, arsenic, beryllium, and nickel caused majority of terrestrial
ecotoxicity. These results are compared below in Figure 4.29, and PHB-Kenaf
composites have twice the impacts for PP-Glass composites. The higher impacts
for PHB-Kenaf composites were due to high electricity consumption, which is

roughly twice as high as the electricity consumption for PP-Glass composites.

177

For PHB-Kenaf composites, following processes made significant
contribution to the total terrestrial ecotoxicity impacts:
- Electricity production (consumed in PHB production process,
composites processing, soybean crushing and cultivation) = 92%
- Steam production (for PHB fermentation and soybean crushing) = 5%
- Miscellaneous steps associated with soybean and kenaf cultivation

(fertilizer production, tractor and truck transport steps) = 3%

 

 

   

CML - Terrestrial EcoToxicity

35” A " '7 H MiPoMnerProducfion

30' IFiberProthdion
2" 25‘ lElectncity-Cormositns
8 Procasing
q 20
:-
u 15~

10--

5

o —

PP-GlouCompooh Ween”

 

 

 

Figure 4.29 Terrestrial EcoToxicity impact for PHB-Kenaf & PP-Glass composites

In case of PP-Glass composites, terrestrial toxicity impacts were largely
due to emissions from Electricity production, and details about processes with
significant contribution to the total impact are mentioned below:

- Electricity production (for composites processing, glass fiber

production) = 98%

- Natural gas production and combustion (glass fiber production) = 1%

178

4.4.2.4 Human Toxicity

Similar to ecotoxicity impacts, electricity consumption contributed towards
the majority of human toxicity for both the systems. Overall, human toxicity
impacts for PHB-Kenaf composites were 64% higher compared to PP-Glass
composites, with air emissions of arsenic, hydrogen fluoride, and benzene being
the main contributor towards this impact. The total impacts for both the systems

are compared below in Figure 4.30.

 

CML - Human Toxicity
7000 ,,

I Polymer Production
' IFiber Production
I Electricity-Conposites
Processing
0 . l . W. WAN . W .77.._._.

PP-Ghu cm PIB-Kenot Commit.

3 ENE? g

S

 

 

 

Figure 4.30 Human Toxicity impacts for PP-Glass and PHB-Kenaf composites

For PP-Glass composites, following processes had significant contribution
towards total impacts:
- Electricity production (for composites processing, glass fiber
production process) = 77%

- Natural gas production (for glass fiber production) = 19%

179

- PP production process = 3%

- Glass fiber production process and raw material production = 2%

Similarly, in case of PHB—Kenaf composites, processes with major

contribution to the total human toxicity impacts are mentioned below:

- Electricity production (consumed for PHB fermentation process,
composites processing, soybean crushing and cultivation) = 88%

- Ammonia production (used as fertilizer for soybean agriculture and as
nutrient in PHB fermentation process) = 6%

- Miscellaneous steps associated with soybean agriculture & crushing,
PHB fermentation (natural gas production, steam production, fertilizer

production) = 6%

4.4.2.5 Depletion of abiotic resources

Abiotic resource depletion impact is determined in antimony (Sb)
equivalents. For the current study, PP-Glass composites involve 55% greater
depletion of abiotic resources, compared to PHB-Kenaf composites (stepwise
comparison shown below in Figure 4.31) and for both the systems, it was due to
consumption of fossil fuels: coal, natural gas, crude oil. The higher abiotic
resource depletion for PP-Glass composites was primarily because of oil and
natural gas consumed for PP production, where it is used as both fuel and

feedstock (propylene production).

180

 

 

CIIL - Depletion of abiotic resources
0.35
l I Polymer Production
0.30 1
F I Fiber Promotion
& 0.25% . . .
o ’ IElectncrthonposm
oE 0.20% I ””59""
E i
g 0.15 4
u l
x .
0.10 i
0.054
0.00 LEA
PP-Ghu Cam Pia-Kenaf Compost

 

 

 

Figure 4.31 Depletion of abiotic resources for PP-Glass and PHB-Kenaf
Composites

The processes, which consumed majority of abiotic resources for PP-
Glass composites were:

- PP production (mainly Oil and Natural gas, used both as fuel and

feedstock) = 72%

- Electricity production (used for composites processing and glass fiber

production) = 21 %

- Natural gas production (for glass fiber production) = 7%

In case of PHB-Kenaf composites, abiotic resources consumption was
mainly because of electricity and steam production processes, with details
mentioned below:

- Electricity production (used for composites processing, PHB production

process, soybean crushing and cultivation processes) = 62%

181

- Steam production (used in PHB production and soybean crushing
processes) = 22%

- Ammonia production (used as nutrient in PHB fermentation and
fertilizer in soybean agriculture) = 5%

- Miscellaneous steps associated with soybean crushing and cultivation
(natural gas production, tractor use, transport by truck, fertilizer

production) = 10%

4.4.2.6 Depletion of the stratospheric ozone

Stratospheric ozone depletion impact (in CFC-11 equivalents) for PHB-
Kenaf composites was 118% higher as compared to PP—Glass composites,
mainly because of their high electricity consumption, which is roughly twice
compared to PP-Glass composites. Air emissions of Halon 1301, methyl
bromide, methyl chloride, and trichloroethane contributed towards this impact for
both systems. A stepwise comparison of the total impact is presented below in
Figure 4.32.

For PP-Glass composites, almost all of the impacts were due to electricity
production process, with equal amounts of electricity consumed for glass fiber
production and composites processing.

Similarly, in case of PHB-Kenaf composites, majority of the impacts were
related to electricity production, with minor contributions from other processes.
The details of these contributions are mentioned below:

- Electricity production (used for composites processing, PHB production

process, soybean crushing and cultivation processes) = 86%

182

- Steam production (used in PHB production and soybean crushing
processes) = 8%
- Miscellaneous steps associated with soybean cultivation (fertilizer

production, tractor use, transport by truck) = 5%

 

CML - Depletion of the stratospheric ozone
4.0E-04 ,

I Polymer Production
3.55-04 ‘
- Fiber Production
3 0E-04
g 2. 5E-04 IElectricity-Corrposites
3 ZOE-04
IL
0
u 1.5E-04
1.0E-04
5.0E-05
0.0E+00

PP-Gllss Collposh PI-B-Ksnsf Composh

 

 

 

Figure 4.32 Depletion of the stratospheric ozone impact for PP-Glass and PHB-
Kenaf composites

4.4.2.7 Climate change

Climate change or global warming impact of greenhouse gases is
computed in C02 equivalents and in the current study, PHB-Kenaf composites
had 14% lower impacts compared to PP-Glass composites. Lower impacts for
PHB-Kenaf composites were mainly due to 002 sequestration by soybean crops,
thus reducing the net 002 emissions of PHB-Kenaf composites. Stepwise
contribution and comparison of PP—Glass and PHB-Kenaf composites to Climate

change impact is presented below in Figure 4.33.

183

 

15000‘

a 602 oq.

10000

 

 

CML - Climate change

I Polymer Production

   
 
  
  

IFiber Promotion
, _- . i 'city _
Processing

PP-Ghu Compoolh FEB-Kenaf Coraposh

 

 

Figure 4.33 Climate change impact for PP-Glass and PHB-Kenaf composites

For PP-Glass composites, following processes contributed towards

majority of climate change impact:

- PP production process = 45%

- Electricity production (used for glass fiber production and composites

processing) = 45%

- Natural gas production and combustion (used for glass fiber

production) = 10%

If C02 sequestration by soybean crops was not considered as a negative

contribution, the total global warming impact for PHB-Kenaf composites would be

57% higher compared to PP-Glass composites, mainly because of C02

emissions related to PHB production process. The details of the contribution by

individual processes are mentioned below:

- Electricity production (primarily used for PHB production process and
composites processing) = 94%

- Soybean cultivation (COz sequestration by soybean crop, emissions
from raw material production and transport steps) = -72%

- PHB production process (C02 emissions from fermentation step and
production of raw materials used in PHB fermentation process) = 68%

- Soybean crushing (emissions from electricity, natural gas, and steam

production) = 8%

4.4.2.8 Eutrophication

Eutrophication impact in the current study was computed in phosphate
equivalents (P043) and production of PHB-Kenaf composites had approximately
six times the nutrifying impact compared to PP-Glass composites, as presented
below in Figure 4.34. Such a high impacts were due to nutrient runoff emissions
(Table 4.18) associated with soybean cultivation, in form of air emissions of NH3
and NOx, water emissions of nitrogenous matter (TN as N) and phosphorus (T P
as P). For PHB-Kenaf composites, following processes contributed towards
majority of eutrophication impacts:

- Soybean cultivation (nutrient runoff) = 89%

- Electricity production (consumed for composites processing and PHB

production process) = 8%

In case of PP-Glass composites, NOx air emissions caused eutrophication

impacts, with major contribution to this impact from following processes:

- PP production process = 47%

185

- Glass fiber production process (NOx emissions from melting & refining,
post forming & finishing steps; from natural gas production and
combustion) = 26%

- Electricity production (used for glass fiber production and composites

processing) = 25%

 

   
 
 

CML - Eutrophication
120 ~
I Polymer Production
100 <
I Fiber Production
9? ,0 .
. ‘ IElectiicity-Compositos
3 Processing
0. 60 “
u
40
20 —.
0
PP-Ghu Coupons Pl-B-Konal Compost

 

 

 

Figure 4.34 Eutrophication impact for PP-Glass and PHB-Kenaf composites

4.4.2.9 Photo-oxidant formation

Photo-oxidant impact (in ethylene equivalents) for PHB-Kenaf composites
was approximately three times (205%) higher as compared to PP-Glass
composites, primarily due to air emissions (of ethylene, CO, methane, and
benzene) related to production of steam and electricity. Comparison of the total

impacts. along with stepwise contribution for PHB-Kenaf and PP-Glass

186

composites is presented below in Figure 4.35. Following processes contributed

towards majority of this impact for PHB-Kenaf composites:

Steam production (consumed in PHB fermentation and soybean
crushing) = 57%

Electricity production (mainly for PHB production and composites
processing) = 17%

Ammonia production (used as fertilizer in soybean cultivation and as
nutrient for PHB fermentation) = 10%

Miscellaneous steps associated with soybean cultivation and crushing
(fertilizer production, tractor use, transport by truck, natural gas

production & combustion) = 16%

 

 

CIIL- Photo-oxidant formation

 

2'5 ; IPolymer Production
l
2.0 J «- IFiber Prouiction
g IElearicityCormosites
£051.51
a 1-0 1

 

PP-Gluo Compost Pl-B-Konof Composh

 

Figure 4.35 Photo-oxidant formation for PP-Glass and PHB—Kenaf composites

187

Similarly, for PP-Glass composites, following processes contributed
towards majority of photo-oxidant formation:
- Natural gas production & combustion (in glass fiber production) = 43%
- Electricity production (used for glass fiber production and composites
processing) = 26%
- PP production process = 17%
- Glass fiber production process (emissions from melting & refining, post

forming & finishing, and raw material production) = 13%

4.4.2.10 Normalization results and Discussion

Final impact assessment results in the current study were normalized, in
order to determine their relative significance, and were computed using CML
normalization factors defined for the World in 1995. This normalization method
was briefly reviewed before in Section 4.4.1.8. The normalized impacts for PHB-
Kenaf and PP-Glass composites are presented below in Table 4.30 and
compared in Figure's 4.36 and 4.37.

Without normalizing the impacts, PHB-Kenaf composites had higher
impacts in 7 out of 9 categories, compared to PP-Glass composites, with lower
impacts for Climate change and Abiotic resource depletion categories. However,
after normalizing the impacts, only following four impacts categories had
significant contribution (Z impacts > 90%) to the cumulative normalized impacts:
Abiotic resource depletion, Acidification, Eutrophication, and Global warming.

Depletion of abiotic resources was greater for PP-Glass composites

mainly because of PP production process, in which 60% of the energy used was

188

as feedstock (crude oil and natural gas). Additionally, their was significant
contribution from composites processing and glass production steps, in the form
of coal consumption for electricity production and natural gas used for glass fiber
production. In comparison, for PHB-Kenaf composites, PHB polymer and Kenaf
fibers were produced using renewable resources and abiotic resource depletion
was mainly due to electricity production and energy requirements for auxiliary
processes, such as production of steam, ammonia, etc.

PP-Glass composites had a higher global warming impact, primarily due
to 002 emissions from PP production process and electricity production; whereas
for PHB-Kenaf composites, global warming impacts were lower compared to PP-
Glass composites because of 002 sequestration by soybean crops, even though
electricity consumption for PHB-Kenaf composites was approximately twice the
consumption for PP-Glass composites. Another observation from global warming
results was regarding the cradle-to-factory-gate scope of the current study.
Because of this scope, C02 absorbed by soybean plants was considered as
sequestered and similarly carbon from feedstock energy (60% of total energy for
PP production) used for PP production was assumed to be sequestered in the
final product. However, there is a difference between the sequestration for
soybeans and PP. If the scope of the current study is changed to cradle-to-grave
analysis i.e. including disposal phase for the composites, then soybean
cultivation process will be considered carbon-neutral, because of the annual
frequency of soybean crops. However, PP LCA will then contain CO; emissions

from feedstock carbon (with incineration as a disposal option), because the

189

feedstock carbon is obtained from non-renewable resources and cannot be
balanced with carbon, which has accumulated over thousands of years.

Acidification impacts were higher for PHB-Kenaf composites, because of
NH3 and NOx emissions from soybean cultivation and high electricity
consumption (thus more NOx and SOx emissions) compared to PP-Glass
composites. The nutrient emissions for soybean cultivation were based on the
second allocation approach specified in a 2005 NREL study4° (Section 4.3.5.1),
which allocates approximately equal nitrate leaching rates for corn and soybean
crops, even though little or no N fertilizer is applied during soybean cultivation.
The original allocation approach in the NREL study allocated nitrogen flows to
the crop to which N fertilizer is applied. However, this allocation approach was
deemed simplistic because of nitrogen fixing by soybean crops (fixed N available
for future corn cultivation) and soil-absorption of excess N left from corn
cultivation (absorbed N available for future soybean cultivation). Because of the
large variation in N-allocation between these two methods, the NREL study
suggested that the most accurate allocation would be between these two
extremes and therefore variation in nutrient emissions and impacts (especially
acidification and eutrophication) using first allocation method is discussed in the
interpretation of the LCA results.

For eutrophication category, as expected, PHB-Kenaf composites had
very high impact compared to PP-Glass composites, because of nutrient
emissions (N and P flows) from soybean agriculture. However, as mentioned

above in the discussion for acidification impact, nutrient emissions from soybean

190

cultivation can vary based on the allocation approach selected and the effect of
such variation on final impacts is discussed in the interpretation of the LCA
results. Additionally, there is significant difference in the air—eutrophication
potentials for CML and TRACI methods (Section 4.4.1.6) and the variation in
eutrophication results by using TRACI methods is discussed in the interpretation
phase.

Summarizing the normalized results, more than 90% of the cumulative
normalized impacts for both systems were attributed to abiotic resource
depletion, acidification, eutrophication, and global warming impacts. Abiotic
resource depletion had the highest contribution to the cumulative impact, with
contribution of more than 50% and 30% for PP-Glass and PHB-Kenaf
composites respectively.

Regarding classification of the normalized impacts, normalization makes it
possible to have a quantitative comparison of the impacts, but not a qualitative
comparison. Methods such as weighting or valuation are usually used to obtain a
qualitative comparison and ISO 1404215 standard recommends that in case of a
comparative LCA study disclosed to the public, weighting should not be used to
draw conclusions about superiority of one product over another. Additionally,

Guine'e et al.19

recommends avoiding weighting if possible, reiterating the
condition set by ISO 14042 standard and in case weighting is performed; using a
weighting method, which is widely accepted and covers all the impact categories.
Since, such a method is still not available, the scope of the current study was

limited to normalization.

191

685. 8.8.282 893 c 885. 85.382 u .28“. a
mmo.:o_mcoE_u 2m womaé 35.25: .8203 fiance .8 3E: ...

 

 

 

88 3.9.8 Ed ”En... «Tu? e: .8 moot} m Eexsoom .5385»
8.3 2-9.» mod «Eta 2.83 e: .8 88:58 a commas“. $58.32.“.
3m: 2-3.» come 8m: Emma .3 .8 3.96 m 5:288 .25 265

:3 3m: mod 2mm; SL3 e: .8 moot. a 3.3.8» 5:5:

mad «tune 3.0 2.8.» Sm; e: .8 N00 a 85:53 .326
8m? 283 3.0 3&3 mimem e: .8 83.4 m Eaxeoom 2.8.2

5o 2mm; mod «En.» 3m? .3 .8 «on. a cowaefioaam

«to 29..» Rd 2-8.8 Emu...” e: .8 Now a 3:85.23

2: «7m: 2: «$3 :83 .3 .8 28:38 9. 5:288 0:22

canon—Eco 0:39:00 moor 6:25
Renown. 3123". .2808 «own-E ants: .3522
885.3 22.83:“.

 

 

 

 

cow—«£35.02

 

 

 

38? .283 E2636 cocozocomv 3%an8 mmflodn. new Ecoxmrn. .2 303:: 35.352 one Each

192

 

Normalized inpacts

“£42
I T I . l E I . 'l
I Phenom Form“
I Ozone layer W
., , {Human Toxicity
3.0E-12 ‘3' I IW .
IAqualic Ecobxicity
I Europhicdion
IAcidificdion
2.05.12 3 Am ' W”
1.0E-12 '
0.05m '-

PHl-Kenal Compost

 

 

 

Figure 4.36 System-wide cumulative normalized impacts

 

 

 

Normalized Impacts
100% . ,, ' * ' ' _

ITerreatrial Ecoloficily
I Photo-oxidant Formoliom
I Human Toxicity
IGlobal Wuming
IAquafic Ecobxiciy
IAcidification
IAbiolic Depletion

m . H,

m .. . z,

0% . , *
PHB-Kenaf Compute "6h“ Cormack

 

Figure 4.37 Normalized impacts: percentage contribution

193

4.5 Life Cvcle Interpretati_9_n_

For a better understanding of life cycle inventory (LCI) and impact
assessment results, ISO 1404321 standard recommends interpretation of these
results using sensitivity analysis, contribution analysis, etc. In the current study,
results were interpreted using contribution and sensitivity analysis. For LCI
results, contribution from individual steps was presented along with total results
and for impact assessment results contribution from individual steps as well as
processes with major contribution were highlighted. Additionally, sensitivity of the
results because of alternative allocation method for soybean farming and TRACI-
Eutrophication method was determined, along with contribution analysis of

energy consumption for both systems.

4.5.1 Sensitivity Analysis - Effect of nitrogen allocation to Soybean cultivation

Soybean cultivation data in the current LCA study is based on a 2005
NREL study“, which quantifies cradle-to-gate emissions and impacts for corn,
soybean farming in the State of Iowa. Since, corn and soybean are rotational
crops, the nutrient emissions due to fertilizer application are related for both
crops. The NREL study follows two different approaches to allocate nitrogen
flows between corn and soybean crops. These allocation approaches were
previously reviewed in Section 4.3.5.1 and are summarized below:

- C-S allocation: approximately equal nitrogen application and leaching

rates for corn and soybean crops

- C-allocation: all nitrogen applied to corn crops is allocated to com

194

For the current LCA study, C-S allocation approach was used to obtain
LCI (life cycle inventory) of PHB-Kenaf composites. As discussed before in
Section 4.4.2.10, NREL study indicates that actual nitrogen flows for corn and
soybeans should lie between these two allocation approaches. Therefore, the
effects of C-allocation on LCI and impacts for PHB-Kenaf composites were
analyzed and compared with the base scenario (C-S allocation) and PP-Glass
composites LCI. C-allocation approach, significantly reduces the amount of FN
(fertilizer nitrogen) allocated to soybean crops and related nutrient outflows of
NOx, N20, TN-SW (total nitrogen to surface water bodies). The modified flows
used for inventory analysis are mentioned below in Table 4.31, all other flows

were same as those specified in Table 4.17 and 4.18 in Section 4.3.5.1.

Table 4.31 Modified-Soybean agriculture average data (1988-2000) "'", allocation
of leaching based on application of N fertilizer (C allocation)

 

Soybean agriculture, C-S rotation, conventional till, no corn stover collection

 

 

 

 

Inputs Units Average (1988-2000)
FN (fertilizer nitrogen) mt-N 4,216
Outputs Units Average (1988-2000)
(a) NOx as N02 mt-N 1,154

(a) N20 as N mt-N 607

(w) TN - SW mt-N 1,782

 

 

 

Lower allocation of FN flows had most significant effect on the

Eutrophication impacts, because of the major contribution (89%) of the nutrient

 

xiii

mt = Metric Tons, prefix (a) signifies air emissions and (w): water emissions

195

flows from soybean cultivation to the total eutrophication impact. And,
eutrophication impacts for PHB-Kenaf composites LCI using C-allocation
approach were 47% lower compared to LCI results using C-S allocation
approach. Overall, using C-allocation approach, PHB-Kenaf composites had
lower impacts for all categories, with small but significant reductions for following
impacts: Abiotic depletion (3%), Global Warming (4%), Acidification (2%), Human
Toxicity (4%), Photo-oxidant formation (6%). For these impacts, except
acidification, lower impacts were due to avoided impacts from lower nitrogen
fertilizer consumption (FN) for C-allocation approach. While, in the case of
acidification, both lower NOx emissions and FN consumption from soybean
cultivation caused a reduction in total impacts.

Regarding comparison of the normalized impacts, using C-S allocation
approach, PHB-Kenaf composites had slightly higher impacts (3%) compared to
PP-Glass composites, as shown before in Figure 4.36. However, because of the
major reduction in eutrophication impacts using C-allocation approach,
cumulative normalized impacts for PHB-Kenaf composites were 9% lower
compared to PP-Glass composites. A comparison of the cumulative normalized
impacts for PHB-Kenaf composites using both allocation approaches and PP-

Glass composites is presented below in Figure 4.38.

196

 

Normalized hnpacts

“5‘" name Ecdoxicity

I Phdcofithnt Formatim

I Ozone layer Depletion

I Human Toxicity

I Glob-l WWW
”5'" manic Ecotoxicity

I Ewoplicstion

IAddficsfion

I Abidic Deplm'on
2E4!
1‘42

-I.‘lE-1|
WCM_ 0-8 mm_ C

 

 

 

Figure 4.38 Effect of variable nutrient allocation (for soybean cultivation) on
cumulative normalized impacts

As evident from the comparison above, based on the allocation method
selected, cumulative normalized impacts for PHB-Kenaf composites, range
between +3% to -9% of the cumulative impacts for PP-Glass composites.
Additionally, the normalization of impacts clearly show that both systems have
same impact categories (Abiotic resource depletion, Acidification, Eutrophication,
and Global warming) making significant contribution to the cumulative normalized
impacts.

Analysis of the energy consumption for both systems and effect of TRACI-

eutrophication potentials on overall results is presented in the following sections.

197

4.5.2 Sensitivity Analysis — Effect of using TRACI Eutrophication impact factors

The CML eutrophication method computes the impacts without
considering location related variation, where such a variation can influence the
actual impact from site-specific emissions. For eg. eutrophication impacts in
CML method are calculated based on the average chemical composition of
aquatic species”: C105H263O110N16P, assuming that a mole of P, N, and COD (as
02) has contribution of 1, 1/16, and 1/138 respectively towards eutrophication.
This assumption is valid for freshwater environments, which are generally P-
limited, but does not hold for marine environments, since they are N-limited.
Additionally, eutrophication results may vary based on the geological profile and
amount of precipitation for a particular location. Therefore, to account for regional
variability and understand its effect on overall results, eutrophication impacts
using TRACI method were calculated, in addition to CML method.

Eutrophication potentials using TRACI are computed by multiplying the
CML eutrophication potentials with US. state-level transport factors to obtain
regional eutrophication factors, and based on these factors U.S. weighted
average values are obtained. These values were previously compared with CML
eutrophication potentials in Section 4.4.1.6, Table 4.26 and present a significantly
lower potentials related to air emissions, while eutrophication potentials for water
emissions are essentially the same. Therefore, TRACI-eutrophication impacts for
PHB-Kenaf composites using C-S allocation and C-allocation were lower by 38%
and 68% respectively. In case of PP-Glass composites, almost all of the impacts

were due to NO)( air-emissions and therefore TRACI eutrophication impacts had

198

even higher percentage-reduction of 85% compared to CML eutrophication
impacts.

Use of TRACI eutrophication factors had an interesting effect on
cumulative normalized impacts. Since, eutrophication impacts had a higher
contribution towards cumulative normalized impacts for PHB-Kenaf composites
(Figure 4.38), reduction in eutrophication impacts using TRACI factors lead to
greater reduction in cumulative impacts for PHB-Kenaf composites compared to
PP-Glass composites, which is evident from the comparison of normalized
impacts shown below in Figure 4.39. As a result, cumulative normalized impacts
for PHB-Kenaf composites using C-S allomtion and C-allocation were lower by

2% and 14% respectively, in comparison to PP-Glass composites.-

 

Normalized Inpacts
435.12
Irma Smithy
I Hub-oxidant Farnah'on

3"" I Ozone hyer Won
I Fluman Toxicity
3542 I Global Warming
IAqntic Ecaoiricily
I TRNSI-Eraropl'ication
LIE-12 IAddficatnn'
IAbiotic Daouion
2.5-12
155-12
1542
05-13
05000 7

WW_ C-8 Wcm_ C PP-Ghaacm

 

 

 

Figure 4.39 Effect of using TRACI impacts on cumulative normalized impacts

199

4.5.3 Analysis of Energy Consumption

Contribution analysis for both systems made it possible to highlight
processes contributing towards the majority of energy consumption and this
information can be useful for manufacturers by helping them focus on particular
processes for energy reduction and environmental improvement initiatives. A
stepwise comparison of energy consumption for PP—Glass and PHB-Kenaf
composites (both C-S and C allocation approaches) is provided below in Figure
4.40. Total energy consumption for PHB-Kenaf composites was 22% (for C-S
allocation) and 25% (for C allocation) lower compared to PP-Glass composites,
primarily because of high energy consumption for PP production process, used
both as feedstock and fuel energy and very low energy consumption for Kenaf
fiber production compared to energy consumed for Glass fiber production.

Even though energy consumption for both systems had strong correlation
to abiotic depletion and global warming impacts, their were differences in
percentage contribution by major processes, with details mentioned below.

For PP-Glass composites, following processes had major contribution to
total energy consumption:

- PP production process (used as feedstock and fuel energy) = 68%

- Electricity production (for Composites processing and Glass fiber

production process) = 24%
- Natural gas production (used in Glass fiber production process) = 8%
In case of C-S allocation for PHB-Kenaf composites, major contribution to

energy consumption was due to following processes:

200

- Electricity production (mainly due to electricity consumed during PHB
production process and Composites processing, with minor
contributions from soybean crushing and cultivation) = 58%

- Steam production (used mainly in PHB production process, with small
contribution from consumption in soybean crushing process) = 16%

- PHB production process (energy consumed for processes such as
cooling water supply and production of raw materials used in PHB
production) = 13%

- Miscellaneous steps related to soybean crushing and cultivation = 8%

- Ammonia production (used as fertilizer in soybean cultivation and

nutrient in PHB fermentation) = 5%

 

I Electricity-Conposites Processing I Fiber Promclion I Polymer Production

Total Primary Energy

700 - , -
600

m .

m

m .

300

200

100 -

0 _.,..._

PP-Gtua Conpoala Pl-B-KanalCompoalL c-s MCmgoalL C

 

 

 

Figure 4.40 Total Energy Consumption for PP-Glass and PHB-Kenaf composites

201

Total energy consumption using C-allocation for PHB-Kenaf composites
was lower compared to C-S allocation because of lower allocation of FN (fertilizer
nitrogen) to soybean cultivation, thereby reducing the energy used for ammonia
production.

Summarizing these results, major energy reductions for the life cycle of
PHB-Kenaf composites are possible by increasing the efficiency of PHB
production process, thus reducing the consumption of electricity, steam, and
other raw materials as mentioned above. Since the technology for PHB
production is still in nascent stage“, increased commercial production will
definitely improve its efficiency. Similar improvements have happened for current
mature technologies, for eg. petroleum and petrochemical industry started in
early 20th century, when coal based industry was dominant; however, advances
in crude extraction technology, basic chemistry (e.g. catalysis), process
technology (e.g. fluidized catalytic cracking) made it possible for petroleum
industry to provide building blocks for the current chemical industry.

In case of PP-Glass composites, production of PP and Glass fibers is
based on a mature technology, thus presenting little opportunity for energy
reduction. However, energy reduction is possible for processing of composites,
because the current calculations for processing composites are based on
production of composites in a laboratory environment (Section 4.3.6) and
industrial production of composites will reduce the energy consumed for

processing composites.

202

References:

 

‘ Sitarz, D., Sustainable America: America’s Environment, Economy and Society in the
21“t Century

2 Netravali, A.N., Chabba, S. Composites get greener. Materials Today, April 2003.

3 Mohanty, A. K., Misra, M., Hinrichsen, G., Biofibres, biodegradable polymers and
biocomposites: An overview. Macromolecular Materials Engineering 276/277, 2000.

‘ wwwgpmeorg. Plastics & Environment: Life cycle & Eco-profiles.

5 ISO 14040: Environmental management - Life cycle assessment — Principles and
framework. First edition, 1997.

6 lntemal communication - Composite Materials and Structures Center.

7 ISO 14041: Environmental management - Life cycle assessment - Goal and scope
definition and inventory analysis. First edition, 1998.

8 Eco-profiles of the European Plastics Industry — Polypropylene (PP). A report by l.
Boustead for PlasticsEurope, March 2005.

9 httpzllwww.eere.energy.gov/industry/glass/analysis.html, Energy and Environmental
Profile of the US. Glass Industry. April 2002.

1° httpzl/wwwmetabolixcoml. Metabolix Brochure, accessed May 2006.

‘1 http://www.biomer.de/, Production of PHB, accessed May 2006.

‘2 Akiyama, M. et al., Environmental life cycle comparison of polyhydroxyalkanoates
produced from renewable carbon resources by bacterial fermentation. Polymer
Degradation and Stability, 80 (2003), 183-194.

‘3 Kim S, Dale BE (2002): Allocation Procedure in Ethanol Production System from Corn
Grain: l. System Expansion. Int J LCA 7 (4) 237—243.

1‘ Kim S, Dale BE (2005): Life Cycle Assessment Study of Biopolymers
(Polyhydroxyalkanoates) Derived from No-Tilled Corn. Int J LCA 10 (3) 200—210.

‘5 ISO 14042: Environmental management — Life cycle assessment — Life cycle impact
assessment. First edition, 2000.

‘6 Bare, J.C., Gloria, T.P. Critical Analysis of the Mathematical Relationships and
Comprehensiveness of Life Cycle Impact Assessment Approaches. Environmental
Science & Technology / Vol. 40, No. 4, 2006.

‘7 Bare, J. C., G. Norris, D. W. Pennington, and T. McKane. 2002. TRACI: The tool for
the reduction and assessment of chemical and other environmental impacts. Journal of
Industrial Ecology 6(3-4): 49—78.

‘8 BEES® 3.0, Building for Environmental and Economic Sustainability, Technical
Manual and User Guide. Barbara C. Lippiatt, Office of Applied Economics, NIST.
October 2002.

‘9 Guine’e, J. B. (Ed.) Handbook on Life Cycle Assessment. Operational Guide to the
ISO Standards; Springer: New York, 2002.

2° Jolliet O et al. (2004). The LCIA Midpoint-damage Framework of the UNEP/SETAC
Life Cycle Initiative. Int J LCA 9 (6) 394 - 404 (2004)

2‘ ISO 14043: Environmental management - Life cycle assessment — Life cycle
interpretation. First edition, 2000.

22 Personal communication, May 2006: Jane Bare, US. EPA - Office of Research and
Development.

23 H.L. Brown, B.B. Hamel, B.A. Hedman, Energy Analysis of 108 Industrial Processes,
Fairmont Press, Liburn, GA 1996.

203

 

2‘ Profile of the Stone, Clay, Glass and Concrete Products Industry, US. Environmental
Protection Agency (EPA), Office of Compliance Sector Notebook Project, September
1995.

25 Glass Industry Scoping Study, Electric Power Research Institute, EPRI EM 59123,
1988.

26 Eco-profiles of the European plastics industry. Report 10: POLYMER CONVERSION.
l. Boustead, May 1997

27 Budget for producing kenaf for year 1999, obtained from Mr. Eugene P. Columbus,
Mississippi State University.

2" Webber, Charles L. 1993. Yield components of five kenaf cultivars. Agronomy Journal,
85: 533-535.

2" Baldwin, B., Kurtz, M., Hoverrnale, O, Neill, S.W., 1996, Kenaf: A guide for production
in Mississippi. MAFES.

3° Stover, J.R., Method for separating Kenaf into core and fiber. United States Patent
Number: 5,970,582. October 1999.

3‘ McKee, H.S., Growing and processing natural renewable BAST fibers for natural fiber
composites. 2"d Annual Automotive Composites Conference, Society of Plastics
Engineers, September 2002.

32 Hansson, P-A. and Mattsson, B. Influence of derived operation specific tractor
emission data on results from an LCI on wheat production. Int. J. of LCA 4, 202-206
(1999).

33 httpzllwww.epa..qov/ttn/chief/gp42/ch09l. section 9.2.2 Pesticide Application, Final
Section-January 1995.

3‘ httpzllwww.dowamocom/usagprodiOM.htm. safety data sheet for Trifluralin herbicide.
35 Sheehan, J. et al., Life Cycle Inventory of Biodiesel and Petroleum Diesel for Use in
an Urban Bus. NREL/SR-580-24089; May 1998.

36 Personal communication with Mr. Brent Brasher, Kengro Corporation.

37 Tillman U. Gemgross, Can biotechnology move us toward a sustainable society?
Nature Biotechnology, Vol 17, June 1999.

3" Tillman U. Gerngross and Steven C. Slater, How green are green plastics? Scientific
American, August 2000.

39 Kurdikar et al., Greenhouse gas profile of a biopolymer, Journal of Industrial Ecology,
4 (2000), 107-122.

4° Powers, Susan E. Quantifying Cradle-to-Farm Gate Life Cycle Impacts Associated
with Fertilizer used for Corn, Soybean, and Stover Production. NREL/TP-510-37500;
May 2005.

‘1 Sheehan, J. Aden, A., Riley, 0., Paustien, K., Killian, K., Brenner, J., Lightle, 0.,
Nelson, R., Walsh, M., Cushman, J., ls ethanol from corn stover sustainable?
Adventures in cyber—farming. Draft Report prepared for the DOE National Renewable
Energy Laboratory, 2002.

‘2 Van Zeijts, H., Leneman, H., Wegener-Sleeswijk, A. Fitting fertilization in LCA:
Allocation to crops and cropping plan. J Cleaner Production, 7: 69-74, 1999.

‘3 Weed, D.A.J., Kanwar, R.S., Water quality - Nitrate and water present in and flowing
from root-zone soil. J Environmental Quality, 25(4): 709-719, 1996.

‘4 Bjomeberg, D.L., Kanwar, R.S., Melvin, S.W., Seasonal changes in flow and nitrate-N
loss from subsurface drains. T ASAE, 39(3): 961-967, 1996.

‘5 Bakhsh, A., Kanwar, R.S., Bailey, T.B., et al., Cropping system effects on NO3-N loss
with subsurface drainage water. T ASAE, 45 (6): 1789-1797, 2002.

‘6 Wang, M.Q., GREET 1.5 - Transportation Fuel-Cycle Model, Volume 2: Appendices
of data and results. Argonne National Laboratory, Argonne, IL, ANUESD-39, 1999.

204

 

‘7 Adams, C. H.; Luttges, W.E.; Mullen, M.; Rand, M.C.; Peters, J.F.; Schroer, B.J.;
Ziemke, M.C. Study on Production, Marketing, and Utilization of Degummed Soybean
Oil as a Diesel Fuel Extender. UAH Report No. 297. University of Alabama in Huntsville,
1981.

‘8 Huijbregts, M. A. J. 1999. Priority assessment of toxic substances in the frame of LCA:
Development and application of the multi-media fate, exposure, and effect model USES-
LCA. IVAM environmental research, University of Amsterdam.

‘9 Guinee, J. B. and R. Heijungs. 1995. A proposal for the definition of resource
equivalency factors for use in product Life-Cycle Assessment. Environ. Toxicol. Chem.
14 (5): 917-925.

5° Wuebbles, DJ, 1988. Relative effects on stratospheric ozone of halogenated
methanes and ethanes of social and industrial interest. UNEP, Nairobi.

5‘ WMO (World Meteorological Organisation), 1992 & 1999. Scientific assessment of
ozone depletion: 1991 8. 1998. Global Ozone Research and Monitoring Project - Report
no. 25 8 44. Geneva.

5’2 IPCC (Intergovernmental Panel on Climate Change). 1996. Climate change 1995: The
science of climate change. Edited by lntergovemmental Panel on Climate Change: J. T.
Houghton, L. G. Meira Filho, B. A. Callander, N. Harris, A. Kattenberg, and K. Maskell.
Cambridge, UK: Cambridge University Press.

53 Heijungs, R., J. Guinee, G. Huppes, R. Lankreijer, H. Udo de Haes, A. Wegener
Sleeswijk, A. Ansems, P. Eggels, R. van Duin, and H. de Goede. 1992. Environmental
life cycle assessment of products. Vol. 1, Guide and Vol. 2, Backgrounds. Leiden, The
Netherlands: Centre of Environmental Science, Leiden University.

5‘ Norris, G. 2002. Impact characterization in the tool for the reduction and assessment
of chemical and other environmental impacts: Methods for acidification, eutrophication,
and ozone formation. Journal of Industrial Ecology 6(3-4): 79—101.

55 Derwent, R.G., M.E. Jenkin, S.M. Saunders & M.J. Pilling, 1998. Photochemical ozone
creation potentials for organic compounds in Northwest Europe calculated with a master
chemical mechanism. Atmospheric Environment, 32. p 2429-2441.

56 Jenkin, M.E. & G.D. Hayman, 1999: Photochemical ozone creation potentials for
oxygenated volatile organic compounds: sensitivity to variations in kinetic and
mechanistic parameters. Atmospheric Environment 33: 1275-1293.

57 Denrvent, R.G., M.E. Jenkin & S.M. Saunders, 1996. Photochemical ozone creation
potentials for a large number of reactive hydrocarbons under European conditions.
Atmospheric Environment 30: 181-199.

5" Huijbregts M.A.J., Breedveld L., Huppes G., De Koning A., Van Oers L. & Suh S.
(2003). Nonnalisation figures for environmental life cycle assessment The Netherlands
(1997/1998), Western Europe (1995) and the world (1990 and 1995), Journal of Cleaner
Production, 11, 737-748. Updated Normalisation factors downloaded from CML website:
httpzllwwwleidenuniv.nl/interfgc/cml/ssp/domains/Icg.html.

59 Stumm, W.; Morgan, J. J. Aquatic chemistry : an introduction emphasizing chemical
equilibria in natural waters; Wiley: New York, 1981.

205

Chapter 5: Biodegradation study of Biobased Composites
5.1 Introduction

It is important to determine the rate of biodegradation of biocomposites
and the effects of disposal end products to the environment. These results can
be very useful in estimating the time to achieve significant levels of
biodegradation and environmental impacts of the end products from composting;
both of the results are valuable in Industry‘, where they are developing biobased
products and will be required to provide these numbers to meet industry
standards as well as federal regulations.

Based on the properties of the material developed there are various end-
of-life disposal options available, such as: composting, incineration, sanitary
landfills, and recycling. Recycling is a feasible option, if the recyclable products
can be easily collected and sent to the recycling facility, where they are
converted to same or new products. For biodegradable materials, disposal option
leveraging upon the property of these materials to degrade makes sense.
Composting and sanitary landfills are established biodegradation options. If the
material developed is non-biodegradable and cannot be recycled, then disposal
by incineration is a viable option, in which energy can be recovered by controlled
combustion of the material.

Sanitary landfills are considered a poor disposal option because of the
way they are designed there is little moisture and microbial activity present, thus
preventing effective biodegradation. Incineration leads to complete disposal,

however incineration facilities are energy intensive and leave toxic residuals.

206

Composting is an aerobic decomposition of organic matter by a mixed
population of microorganisms under controlled conditionsz. Compost material
behaves as a porous material with moisture content of 50 to 60 percent and
through which air can be circulated to supply 02 and to remove evaporated H20.
Composting is an environment friendly approach to convert biodegradable waste
including biodegradable plastics into useful soil enhancement products3 as
discussed in more detail in section 2.c.

It is clear that composting has significant advantages over other disposal
modes for biodegradable products. Therefore, it was selected as the way to
determine the rates of biodegradation and environmental impacts of disposal end
products.

ASTM and ISO standards relevant to demonstrating compostability of
plastic materials have been discussed in section 2.c.ll. ASTM D 6400 and D
6002 do not have equivalent ISO standards, and since ASTM standards provide
greater detail about experimental set-up and related calculations, they formed the
basis of this experimental study. ASTM D 5338 provides lot of details about
setting up a controlled composting apparatus, and is reviewed in the following

section on experimental set-up and procedure.

207

werimental set-Hp ancflrocedjgrg

ASTM standard D 5338 defines the percentage of biodegradability as
obtained by determining the percentage of carbon in the test material that is
converted to 002 during the duration of the test. The percentage biodegradation
does not include the amount of carbon converted from the test substance to cell
biomass, which in turn is not converted to C02 during the test. In the past there
have been studies done, where the complete carbon balance for the composting
vessel was accounted for by using an inert solid medium with all necessary
nutrients for microbial growth and inoculated with aqueous compost eluate‘,5,6.
However, such a study is very time consuming and is beyond the scope of D
5338, which is intended to provide rapid and reproducible determination of
biodegradability of the test substance.

The standard described above recommends two options for experimental
set-up in order to capture CO; from controlled composting experiment. In the first
option 002 can be captured using a carbon-dioxide scrubbing solution such as
Ba(OH)2 and regular determination of CO; is made by titrating the scrubbing
solution with HCI.

The schematic in figure 5.1 describes the set-up required for testing only
one sample by this option. The standard requires users to test samples in
triplicates in a composting vessel volume of 2 to 5 Liters. Other requirements are:

I. Controlled temperature conditions of 58°C (12°C), which can be maintained

by using a water bath or other temperature controlling apparatus such as

convection or radiation heaters.

208

II. A pressurized air system, which provides CO; free, water saturated air to
each composting vessel at accurate flow rates. If direct determination of
exhaust air is being done (further explained in second option) then normal
air may be used.

Ill. Suitable device such as sensors, gas chromatograph to measure the O;
and CO; concentrations in the exhaust air of composting vessels.

IV. Recommended carbon dioxide trapping apparatus for individual composting
vessel should be combination of three 5 Liter bottles fitted with gas sparging
and containing Ba(OH); carbon-dioxide scrubbing solution. End of
experiment titration with 0.05 N HCI provides the amount of CO; produced
by the test substance.

Based on these requirements, conducting the controlled composting study
by this option requires a large amount of laboratory space and a controlled
temperature environment large enough to hold a minimum of 12 composting
vessels required for testing the biodegradability of one sample.

The second option for conducting the experiment is by flow measurement
of exhaust gas from composting vessels and using a gas detection apparatus
such as a gas chromatograph or CO;/O; sensors to measure the CO; and 0;
concentrations for each vessel. An example of such a set-up is provided in figure

5.2 (reproduced from ASTM D 5338 standard).

209

Air (free of 00;)

B = Blank
P = Positive Control

N = Negative Control

G = Test Specimen
G) = Scrubbing solution Ba(OH)2

III?

Figure 5.1 Experimental set-up using Carbon-dioxide trapping apparatus

210

CONTROLLED-COMPOSTING BIODEGRADABILITY

.'.'.‘.¢
5
3.
:1 .
‘ -: a PRINTER
'J

DATA
PROC ESSOR

 

 

VIII’eI-uoo-v
' ."I.oo-oo.l
'.

 

 

 

 

f1

 

 

 

‘° {22:23——

 

 

 

 

 

 

 

55 :3 . "6 .. GAS
-: \g _ .3 CHROMATOGRAPH
:1 t g
35 if
5 05
:’ :’
:f' r; 1 Air 6 Multiport valve
:3 :3 2 Flow Control 7 Test Gas

_ . 3". "w. 3 Humidifier 8 Carrier Gas
“ """"" 4 Composting Vessel 9 Sample Loop

INCUBATOR COOLING UNIT 5 Condensate 10 Flow meter

Figure 5.2 Experimental set-up using Gas Chromatograph (reproduced from
ASTM D 5338 - 98)

In order to determine the rate of biodegradation by this second option,
ASTM D 5338 recommends analyzing the CO; concentration on a frequent basis
during the test period (frequency of 3 to 6 hours is provided as an example in this
standard). Frequent CO; analysis ensures the reliability of the cumulative CO;
production numbers obtained over the course of the test.

Experimental set-up for the current study was constructed similar to
second option and is described in figure 5.3 (details in figure 5.3.1 through 6).
Figure 5.3.1 describes the set-up used to obtain clean air at a reduced pressure.
Laboratory air was available; however because of the high pressure in the range

of 100-120 psi and excess moisture and oil

211

Flow
direction

 

Pressure regulator Air Filter Desiccant-Air dryer

Figure 5.3.1 Pressure Regulator and Air
treatment

Flow meter/
controllers

Thermometer

Temperature
controller

 

Air Process Heater

Humidifying Flask

Needle valves

Flow distribution
Manifold

 

Figure 5.3.3 Heater set-up and Flow distribution

212

 

Figure 5.3.4 Composting vessel

Desiccant
Air dryer

Sampling port

 

Figure 5.3.5 Exhaust air collection set-up

 

Figure 5.3.6 Gas Chromatograph with TCD detector

213

    

Figure 5.3 Overview of the Composting Study - Experimental Set-Up

present (because a refrigerated air dryer was not installed), pressure regulator
(Precision aluminum stacked filter/regulator 0 to 10 psi pressure, ‘/4" pipe, 22
SCF M Max, ControlAir lnc., NH USA) coupled with air filter (Speedaire® Air Line
Filter, ‘A inches, Dayton Electric Mfg. 00., IL USA) and desiccator (Excelon 74
Desiccant Compressed Air Dryer, 1/4” pipe, 20 SCFM Max, 100 psi, Norgren, CO
USA) were used to obtain clean-dry air at a reduced pressure ranging from 2-6
psi.

Air obtained from the set-up described in Figure 5.3.1 was bifurcated
using a barbed T-connection and distributed through two flowmeters (Model
VFB-55-BV, Range 20 — 200 SCFH Air (ft3/hr), Dwyer Instruments, Inc., IN
USA) to Glove box A and B as shown in Figure 5.3. Air-flow to the glove boxes
was recorded in the range of 50 — 60 SCF H and acted as an indicator to identify

any flow obstruction and also to ensure that minimum flow requirement for the Air

214

process heaters was met. The temperature inside the glove box was controlled
using a dedicated benchtop temperature controller (Monogram Benchtop
Controller, Model CSC32T using T type thermocouple, Omega Engineering, Inc.,
CT USA). Assuming a cuboid-shape for the glove box, a thermocouple was
suspended to get a reading from the approximate center of the glove box and
connected to the benchtop temperature controller so that temperature inside the
glove box was controlled based on the average temperature of the glove box.
Set-up for flow and temperature control is illustrated in figure 5.3.2.

Clean dry air was required for the experiment, so as to avoid any
contamination due to excess oil and water in the compressed air, and to meet the
requirements for operating air process heater used for this study. Air Process
heater (Model AHP 7561, ln-Line Air and Gas Heater, 750 Watts, Omega
Engineering, Inc., CT USA) used for the study requires clean, dry incoming air,
in order to avoid heater filament burn out. Wattage requirement for the glove box
was calculated based on the specific heat of compost and test materials and
estimated as 815 Watts for heating 15 composting vessels in a glove box to 58°C
in one hour, detailed calculation is provided in Appendix 5.2.1.

As specified in Appendix 5.2.1, exit air temperature from the air process
heater can reach up to 250°C. Copper tubing with ‘/4” OD and 0.030” wall
thickness was used to transfer air from heater to a modified 2 L Erlenmeyer flask,
acting as a humidifying vessel for the experiment. The Erlenmeyer flask was
modified with 1A” glass-to-metal union made at 700 mL gradation of the flask, as

shown in figure 5.4. The flask was modified in order to avoid the gradual

215

degradation of the rubber stopper from exposure to the high temperatures (in
excess of 150°C) of the copper tubing transferring heated air from the air process

heater.

 

Figure 5.4 Modified 2 Liter Erlenmeyer flask

To ensure uniform temperature distribution and eliminate any cold-spots in
the glove box, two circulating fans (3" AC Fan, Radioshack) were mounted in the
glovebox.

Since, ASTM standard D 5338 requires the incoming air to the composting
vessels to be water saturated, two humidifying flasks were used in succession. A
total of 13 composting vessels were tested in one glove box. Heated and water
saturated air from the second humidifying flask was bifurcated and transferred to
a 15 outlet, Nylon Manifold Block (3/4" NPT Inlet x 3/8" NPT Outlet, McMaster-

Carr, OH, USA) using high temperature rubber tubing. Two outlets at the

216

extremes of the manifold were closed using plugs, since the experiment required
only 13 outlet connections.

A manifold was used to create an ideal flow condition with equal flow rates
to individual composting vessels. Previous controlled composting studies
provided a good starting point to optimize the incoming flow rate to individual
composting vessels. Gattin et al.5 have used a flow rate of 500 mUmin in the
study to calculate the mineralization of starch in compost media. While,
Jayasekara et al.7 in their study to evaluate biodegradability for cellulose,
newspaper and starch-based polymers used a flow rate of 166 mUmin for each 3
Liter composting vessel. In the current study, based on the preliminary
experiments with compost and positive control (compost + cellulose), flow rates
in the range of 100 mL/min, lead to CO; detection in the range of 0-6% using a
gas chromatograph with a TCD detector. However, direct flow of air from the
manifold outlets was too high to obtain low flows of ~100 mUmin; therefore
needle valves (F200B flow control, 11 LPM Max Flow, Parker Hannifin
Corporation, OH, USA) were used to reduce the flow rate of air into composting
vessels. A nylon manifold block along with needle valves was mounted on a
wooden board, with push-connect fittings used on both ends of needle valves to
connect to manifold as well as composting vessels. Heater set-up and flow
distribution have been described in figure 5.3.3.

Sufficient care was taken to properly imbed the incoming air tubing into
composting mix, in order to create an aerobic environment for composting and

have uniform heat transfer. Regular temperature readings of the composting

217

vessels were taken using the imbedded T-type thermocouple in composting
vessels; so as to ensure that temperature of around 58°C were maintained in the
composting vessels as per ASTM D 5338 standard. An example of the
composting vessel is shown in figure 5.3.4.

Moisture removal and cooling to room temperature of the air is necessary
in order to get an accurate reading using Gas Chromatograph. Exhaust air from
the composting vessel was cooled to room temperature and dried, before
collecting the sample for Gas Chromatograph analysis. Collection set-up used for
the experiment is described in figure 5.3.5. Coiled-Copper tubing (W OD and
0.030” wall thickness) was used to cool the exhaust air from composting vessels,
before drying it using a Drierite drying tube (Fisher Scientific, USA).

Sample collection consisted of collecting sample from 125 mL Erlenmeyer
flask; however because of the possibility of contamination due to inflow of air
during sample drawing, final set-up was modified to collect sample directly from
flexible tubing using removable Swagelok® fitting, fitted with 5” long Iuer hub-22
gauge sideport needle. Sideport needle was passed through silicon-septa in
order to create a gas tight environment for the gas collection apparatus. Modified
set-up is shown in figure 5.5.

Gas samples were collected using 1 mL SGE GT (Gas-Tight) Syringe with
Retro-fit SLLV—syringe valve (both syringe and valve purchased from Alltech
Associates, Inc., IL, USA). SLLV-syringe valve is used to ensure a gas-tight
environment, and to release excess pressure and equilibrate to atmospheric

pressure in case of a pressurized sample. Since the composition of gas sample

218

measured by gas chromatograph is in relative terms of percentage of particular
component (say CO2) in the total sample, pressurized gas sample can
significantly increase the percentage reading and thus produce error in the actual
gas composition reading. Assuming an exhaust gas flow rate of 60 mL/min,
flushing time of two minutes and sampling time of four minutes was found to be
sufficient after connecting the exhaust from composting vessel to the collection

apparatus.

 

Figure 5.5 Modified exhaust gas collection set-up

Samples from the collection apparatus were manually injected (using 1 mL
SGE GT syringe fitted with Luer Precision Side-Port Needle 0.028” OD x 0.016"
ID x 2" Long) into an on-column injector on SRI-Gas Chromatograph (Model
3100, SRI Instruments, CA, USA) with TCD (Thermal Conductivity Detector)
detector, as shown in figure 5.3.6. A packed column (packing: PORAPAK Q
80/100, 6’ x 1/8” SS column, purchased from Alltech Associates, Inc., IL, USA)
with ability to separate Air/O; from CO; was used to determine the CO;

concentration of the injected sample. The gas chromatograph was connected to

219

a computer and data was collected using chromatography integration software,
PeakSimpIe 2.83 (SRI Instruments, CA, USA).

For sample analysis and developing standard calibration curve, following
GC conditions were used:

Career Gas: Helium, 99.995% purity, <O.5 ppm THC (Total Hydrocarbons)

Carrier Gas Pressure: 11 psi

Oven Temperature: 40°C

Standard Calibration Gases: Pure Carbon Dioxide 99.8%; Multicomponent
gas mixture, 3% CO2, 17% O; in N; (both MicroMAT 14 L, purchased from
Alltech Associates, Inc., IL, USA); Scotty Gas Mix 1% CO;, 20% O; in N; (14 L,
Sigma-Aldrich, PA, USA)

A brief review of the concepts of Gas Chromatography and reasoning
behind using TCD detector for the current study were presented earlier.

Based on the preliminary degradation experiments, CO; calibration curve
for the concentration range of 20% - 0.2% CO; was developed. Additionally for
greater precision in measuring low concentration readings, another calibration
curve with a range of 3% - 0.2% CO; was also developed. Calibration data (table
5.1) was collected using 1 mL SGE GT syringe; the same kind of syringe was
also used for sample collection and injection. Calibration curves along with linear

equations and correlation coefficients are shown in figures 5.6 and 5.7.

220

Table 5.1 CO; calibration data

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Standard Calibration Sample C02 Retention time
Concentration Area
Gas type size (mL) (minutes)
(%)
99.8 % CO; 0.20 19.96 1.083 185.82
99.8 % CO; 0.14 13.97 1.116 117.56
99.8 % CO; 0.10 9.98 1.116 64.99
3% CO;, 17% O; * 1.00 2.95 1.166 33.55
3% CO;, 17% O; * 0.50 1.48 1.166 17.29
3% CO;, 17% O; * 0.30 0.89 1.183 9.87
1% CO;, 20% O; ** 1.00 1.02 1.183 11.57
1% CO;, 20% O; ** 0.50 0.51 1.183 5.62
1% CO;, 20% O; ** 0.20 0.20 1.183 2.09

 

*actual standard CO; concentration = 2.952 %
"actual standard CO; concentration = 1.02 %

221

 

 

Calibration curve for CO; GC Analysis

 

 

 

 

i (20% ' 01% C02) y = 8.6522x + 0.7436
3 R2 = 0.9786
I 200 ------smmm -----W__M__._-._-_-- I
I 150 1
N F—
l 2 100 - (+393 I
j < L— Linear (Area) I
I 50 ~
‘ 0 I T
I 0 5 10 15 20

I Concentration (%)

... __

Figure 5.6 CO; calibration curve (20% - 0.2%)

 

77777 *“I

 

 

 

._.-4___ # - fl _i.—A_777‘___#_fi4

 

I Calibration curve for CO; GC Analysis
'| (3% - 0.2% CO;)

40
35 -
30 ~
25 -
20 ~

 

Area

10~

 

 

 

0 0.5 1 1.5 2 2.5 3
Concentration (%)

 

y =11.453x - 0.105
R2 = 0.9996

 

F—o— Area
15 - 1— Linear (Area)!

I

 

 

Figure 5.7 CO; calibration curve (3% - 0.2%)

222

 

 

5.4 Compost Characterizatigfl

Compost used for the biodegradation study was characterized, in order to
determine its suitability for the study as required by ASTM D 5338 standard. The
standard recommends that compost used should be two to four months old,
obtained from municipal solid waste or plants or yard waste or a mixture of either
of them. Additionally the compost should have ash content of less than 70%, pH
between 7 and 8.2, and total dry solids between 50 and 55%. The compost
inoculum should be free from inert substances such as glass, stones, metals,
etc., and should be screened in order to remove such material from the compost
received.

In this study, plant and yard waste based compost was used and had
following characteristics:

Total dry solids = 60.08%

Volatile solids = 61.74%

pH=79

% Nitrogen = 1.8, cm = 16:1

Percent Nitrogen of the compost was determined at Soil and Plant
Nutrient Laboratory (Dept. of Crop and Soil Sciences, Michigan State University)
using TMECC 04.02-D8 using N-Analyzer and following Dumas Method.

In addition to compost characteristics above, the maturity of the compost
was determined using the Solvita® procedure9 and the Maturity Index of the
compost sample. The maturity index measurement was carried out by detecting

carbon dioxide and ammonia levels in the compost, in order to assess the

223

maturity, which is defined in the Solvita® procedure as “resistance to further
decomposition and absence of phytotoxic compounds“.

For the Solvita® test, compost was screened to remove inert materials and
obtain uniform size compost and the moisture content was adjusted to create
ideal composting conditions. This compost was loaded into standard solvita jar
(without compression) to the indicated line. Maturity index is a two-dimensional
index, which depends on CO; and ammonia measurement. Therefore, two
separate standard solvita jars were used to measure compost-CO; and ammonia
emissions using chemically sensitive gel paddles. These paddles have different
initial color (CO; paddle is purple, Ammonia paddle is yellow) and based on the
CO; and ammonia emissions (over the test period of four hours in the closed jar)
change color. These changes are compared to the standard scales, in order to
determine the CO; and ammonia indices, which in turn provide an overall
Compost maturity index as described in Figure 5.8. Interpretation of the compost
maturity index is provided in Table 5.2.

Compost in the current study was similarly tested for 4 hours for CO; and
ammonia emissions and at the end of the test; paddles were removed from the
jar and compared to the standard color chart. These comparisons are provided in
Figures 5.9 and 5.10. With carbon dioxide test result of 7 and ammonia test
result of 5, compost maturity index was 7, which corresponds to well matured,

aged compost with few limitations for usage, as described in Table 5.2.

224

€833.00; new 05.0336
E_>_ow _m_oEO Eat noosuoaocv 2an x09: 3......sz Fmanoo ucm mcfiow EmucoFm m_:oEE< .23on .5960 Qm 0.59“.

1.3.852?! 2: .6... .o I. =32 Nco a... ...... .N a. :32 £2 2:. insane ...

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

e N N F F F F F £23.. .95 F m
... m s F N F F F ...z .....F. N ...-m.
N .. m s N N F F £25.32 N. a.“
a F e m s n N F £2 33 F. m
a N e ... s m N F £sz 80> m . m.

0.

w n ... m w m N _
..im SFFIQK ~I.Fl—. AJZZGFC ...-1:59 (..r—LFVQCV. .

Foo:— ufiaioo I5 82.: o. 5.6... o:- Ioohoo to! ...—Fe 238:... 3.3 £33.. N00 oco 63.5.55 05 «I:

.63.? 8....— EFBNE Fax—Eco ._ 2...:

   

.. m

.. a m

m

o I m
6.85:2 8... . . FN..rroo ”.9: ’50 26..

. . . a: comm—8.059.. ..8 331 .
asifiuufli again; ...: .512. «My. 33!: afifigfifizd.) 2
. .8882: hmMP mD-XOE . ZOQm—(O

memhfiz. Om 3.qu o<H F >AO l E55. $9.38 .<._. :50 B

225

Table 5.2 Interpretation of Compost Maturity Index (adapted from Official Solvita®
Guideline)

 

Composting
Corresponding Composting Process:
Maturity Index

 

Inactive, highly matured

compost Finished Compost

 

7 Well matured, aged compost

 

Curing; aeration requirement
6 Curing
reduced for the compost

 

Compost moving past the

 

 

 

 

5 active phase of
decomposition and ready for
curing Active Compost
Compost in medium or
4 moderately active stage of
decomposition
3 Active Compost Very Active Compost
2 Very active; putrescible fresh
compost; high respiration rate Raw Compost
1 Fresh, raw compost

 

 

 

 

 

226

 

COMPARISON
Figure 5.9 Solvita Carbon Dioxide Test Result

   

START

 

COMPARISON
Figure 5.10 Solvita Ammonia Test Result

227

5.5 Singles tested: processing and characterization
A list of the materials, whose biodegradability was tested, is provided

below in Table 5.3.

Table 5.3 Materials tested for Biodegradability

 

 

 

No. of
Test Substance Type of material
samples
Compost Aged, mature compost 3
Cellulose-positive
Microcrystalline Cellulose”, 13 pm particle size 3

control

 

PP — (30wt°/o)
Injection molded composite samples of PP11
Glass-negative 3
(Polypropylene) and 1/4 " Textile Glass fibers12
control

 

PHB - (30wt°/o) Injection molded composite samples of PHB13

f14

Kenaf and ‘/4 " Kena fibers

 

PHB — (5wt%)PHB- Injection molded composite samples of PHB
g-MA + (30wt°/o) and 1A " Kenaf fibers with maleated PHB added 3

Kenaf as a compatibilizer

 

Injection molded samples of PHB
PHB 1
(Polyhydroxybutyrate) polymer

 

Injection molded samples of PLA15 (Poly-lactic
PLA 1
acid) polymer

 

Injection molded samples of CAT6 (Cellulose
CA - (30wt%) TEC 1
Acetate) plasticized with TEC” (Triethyl Citrate)

 

 

 

 

 

228

Microcrystalline cellulose powder10 with a particle size of 13 pm was
selected as a positive control, since under standard composting conditions it is
biodegradable and additionally it satisfies the ASTM D 5338 criteria of selecting
analytical-grade cellulose with a particle size of less than 20 pm.

All the composites and polymer samples except PLA were initially
extruded using twin-screw extruder ZSK 30 (Werner Pfleiderer), pelletized and
then injection molded into dogbone shape, standard tensile coupons using 85
ton-Cincinnati Milacron injection molder. PLA polymer was available as pellets;
therefore tensile coupons were directly obtained by injection molding.

As recommended by ASTM D 5338; dry solids, volatile solids, carbon and
nitrogen content data was determined for all the samples and provided in Table
5.4 below.

Dry solids and volatile solids content of the samples were determined
using APHA (American Public Health Association) standard 2540 G-Total, fixed,
and volatile solids in solid and semisolid samples”. For calculating dry solids
content, samples were dried in a convection oven at 103 to 105°C, until weight
change was less than 4% or 50 mg, whichever was less. For the volatile solids
test, the dried residue samples from dry solids test were heated in a Muffle
Furnace (T hermolyne — Type 30400, Automatic Furnace) to 550°C, until weight
change was less than 4% or 50 mg, whichever was less. Carbon and Nitrogen
content of the samples was determined using Carlo Erba NA15 Series 2, CNS
Analyzer (Plant, Soil, and Carbon-Nitrogen Analysis Laboratory, Dept. of Crop

and Soil Sciences, Michigan State University).

229

For a good composting process, the cm ratio of the mixture of compost
and test substance should be between 10 and 40, as recommended by ASTM D
5338. The C:N ratio for the compost-sample mix (see Table 5.4 below) was in the

recommended range between 10 and 40.

230

Table 5.4 Sample Characterization

 

 

 

 

 

 

 

 

TEC

 

 

 

 

 

 

Test Substance Dry Volatile Carbon Nitrogen Compost+
Solids Solids Content Content Sample
content content (%) (%) C:N
(%) (%)
Cellulose-
96.10 95.34 42.07 0.00 20:1
positive control
PP — (30wt%)
Glass-negative 99.94 74.47 78.26 0.00 23:1
control
PHB - (30wt%)
98.83 98.35 52.67 0.00 21:1
Kenaf
PHB —
(5wt%)PHB-g-
98.90 97.99 51.97 0.00 21:1
MA + (30wt%)
Kenaf
PHB * 99.92 54.10 0.00 21:1
PLA * 100.00 46.70 0.00 20:1
CA — (30wt%)
* 97.18 46.60 0.00 20:1

 

 

' Not determined, but assumed to be 100% because of neat polymer being used

231

 

5.6 Cmtrplleg composting experiment: start-up, operating procedure, and
results

5.6.1 Start-up

Main objective of this experiment was to simulate the actual composting
conditions in a laboratory environment and study the degradation behavior of the
polymers and their composites listed in Table 5.3. And, optimal composting
conditions exist with moisture content/dry solids of 50% and temperature of 58°C
(12°C). As discussed before in experimental set-up, temperature was maintained
around 58°C using a combination of temperature controller, air-process heater
and circulation fans in the glove box. Based on the dry solids content of compost
and test samples, moisture content was adjusted using distilled water to 50% by
weight.

All the test materials (polymers and composites, except positive control
and compost) were in the shape of standard dogbone-tensile coupons.
Therefore, their size was reduced to less than 2 by 2 cm, so as to increase the
available surface area for degradation and better mixing of samples with
compost. Cellulose positive control was available in powder form with a particle
size of 13 um. Mixing of test material with compost and addition of moisture leads
to significant reduction in the porosity of the mixture; therefore an inert mineral-
material called Perlite19 was used to increase the porosity of the mixture. Perlite
was obtained from The Scotts® Company (OH, USA) and 400 mL or less (based
on the density of the mixture) was well mixed with the compost-test material

mixture.

232

After sufficient mixing has been obtained, the mixture was loaded into 2-
Liter Erlenmeyer flasks. Proper precautions were taken to avoid filling up the test
material upto the top of the flask, which can lead to insufficient headspace for
gas collection and periodic manual mixing of the test sample. As a rule of thumb,
samples were filled upto three quarters of the volume of 2-L Erlenmeyer flask.

These loaded Erlenmeyer flasks were closed with size 10, 3-hole Twistit
rubber stoppers (VWR International, IL, USA), and sealed with silicone sealant to
ensure gas-tight environment. One of the three holes was used to embed the
incoming hot-air line into the compost, another for embedding the T-type
thermocouple in the compost for regular temperature readings, and third hole for
exhaust gas collection from the headspace above the sample in the flask.

Before connecting the loaded and sealed samples to the flow manifold,
these samples were weighed to record the initial weight at the beginning of the
experiment. This reading in combination with final weight reading at the end of
the experiment was useful in determining the weight loss and amount of sample

converted to carbon dioxide.

5.6.2 Operating Procedure

Since composting is an aerobic degradation process, proper aeration in
the composting vessels was maintained. ASTM D 5338 standard recommends
that oxygen levels should not drop below 6% in the exhaust air. In the current

experiment maximum 002 level of 13.4% was recorded, therefore based on the

233

original composition of air”, oxygen levels greater than 6% were present in the
composting vessels at all times during the degradation study.

As described in section 5.2, temperature in the glove box was controlled
using temperature controller, which controlled the operation of air process heater
and additionally circulation fans were used to have uniform heating in the glove
box. Daily temperature readings were recorded from individual composting
vessels using Fluke digital thermometer. These readings provided a good
indication of the status of the composting process. For example, if the
temperature reading was above 58 :l: 2°C, it indicated either hyperactive
composting conditions or low flowrates. In case of temperatures below the
above-mentioned range it was due to retarded composting conditions or
extremely high flowrates. Based on the CO; concentration readings using gas
chromatograph and temperature readings, optimal composting conditions were
maintained by either adjusting the incoming air flowrate or changing the set point
for temperature controller.

Two glove boxes (A&B) were used for this study (see figure 5.3) and
samples loaded were named based on their position in the glove box and
suffixed with glove box name (A or B). Rate of biodegradation of the samples
was determined using gas chromatography analysis of the effluent gas from
composting vessels on regular basis. In section 5.2 (experimental set-up and

procedure), I have discussed two different ASTM recommended methods to

 

‘ Dry air composition at sea level is: nitrogen 78.08%, oxygen 20.95%, argon 0.93%,
carbon dioxide 0.03%, neon 0,0018%, helium 0.0005%, krypton 0.0001%, and xenon
0.00001 %. Reference: "Earth’s atmosphere” A Dictionary of Chemistry. Oxford
University Press, 2000. Oxford Reference Online.

234

measure biodegradability: titration, gas chromatography. Gas chromatograph
method was selected because it requires less bench space as compared to
titration method. Average gas sampling from samples in glove box A was 1.1
days, and for glove box B was 1.3 days.

Proper aeration and moisture content was maintained in the composting
vessels, for optimal composting conditions during the experiment. The complete
aeration system for the glove box was checked for leaks regularly. If any
composting vessel had excessively dry conditions, additional water was added to
increase the moisture content. In addition to these specific adjustments, every
two weeks, the whole system was shut down in order to fix any leaks, add
moisture and manual shaking of all the composting vessels.

ASTM D 5338 standard recommends 45 day controlled composting study,
which can be extended in case significant biodegradation is still being observed.
In the current research, biodegradation study was conducted for 60 days,
because of time constraints and fulfilling the project objective of determining the

rate of biodegradation of biocomposites.

5.6.3 Results

Concentration of 002 (%) in the exhaust gas from composting vessels
was measured by a gas chromatograph and converted to mass basis using ideal
gas law equation and flow rate measurement of the exhaust gas. An example
calculation is provided in Appendix 5.6.3.

As described in Appendix 5.6.3, %age C02 detected in the exhaust of

composting vessels was converted to grams of C02 per reading and grams of

235

carbon per reading. Biodegradation progress for all the samples is described in
time plots by displaying cumulative carbon dioxide evolution over time. The
percent biodegradation for all the samples was calculated using the formula

mentioned in ASTM D 5338 standard:

mean Cg(sample) - mean Cg(blank)

x100
Ci

% biodegradation =

 

where:

C9 = amount of carbon produced in biodegradation, grams

C; = total amount of carbon in the test sample, grams

Additionally, standard error, 8., and confidence interval (CI) of the
biodegradation readings were calculated to determine the variability of the results
obtained.

Standard error (8.) is defined as:

 

2 2
S s 10
S = sample blank
° J“ n1 )+( n2 ”x ci

where n1, n2 are number of test sample and compost control samples
respectively; s is standard deviation for carbon production during biodegradation,
and Ci is the total amount of carbon in the test sample.

Confidence interval (CI) is defined as:

x% CI = %biodegradation :l: (t x Se)

where t is the t-distribution value for x% Cl with (n1+n2-2) degrees of freedom.

236

5.6.3.1 Results-Compost Control

A cumulative 002 evolution curve for compost control samples is shown in

Figure 5.11 below.

Cumulative COZ_Com post Control

70 .1 - ..z.--__-.-m. ”WWW“--- .._.__ --__

497); ' /

 

8

 

OD #-
O 0

Cumulative C02 (grams)
N
O

 

 

.t
C

 

0 r r r r r r
0 10 20 30 Days 40 50 60

Figure 5.11 002 evolution time plot for Compost Control

Composting study was started with compost control samples in triplicates;
however one of the samples was rejected because of non-ideal conditions during
the study and thus difficulty in obtaining reliable readings. C02 evolution curve for
the other two samples indicate a similar rate, which by the end of the experiment
at 60 days achieves similar amount of 002 being produced, even though the

samples are in separate glove boxes (glove box A&B). This result indicates a

237

good quality control for this study. Cumulative amount of carbon produced during

the study for these samples is provided in table 5.5 below.

Table.5.5 Cumulative amount of carbon produced-Compost Control

 

 

 

 

Sample ID Cumulative carbon produced (grams)
7A 17.53
8B 17.72
Mean :l: Standard Deviation 17.63 i 0.13

 

 

 

 

5.6.3.2 Results-Cellulose positive control

Cumulative C02 production time plot for the cellulose positive control-
compost mix and compost control sample-7A is presented below in Figure 5.12.
Compost control sample was included in the time plot, in order to provide a C02
evolution comparison between cellulose and compost control samples. One of
the three cellulose positive control samples was rejected because of non-ideal
conditions. For other two positive control samples CD; production during the
composting study was very similar; therefore similar to compost control results, it
indicates good quality control for the experiment. Cumulative carbon production
and percentage-biodegradation numbers for cellulose-positive control samples

are reported in Table 5.6, below.

238

Cumulative COZ_Cellulose-positive control

 

+ 4B
+ 8A
“ + compost blank

 

..t

O)

O
1

 

_L
h
C

 

 

..A
N
0

Cumulative 002 (grams)
8 §

03
O

 

A
O

 
    

 

 

T T f T T I

10 20 30 Days 40 50 60

N
O O
l

C)

Figure 5.12 CO; evolution time plot for Cellulose positive control

In addition to percent-biodegradation, standard error and 95% confidence
interval was calculated of the percent-biodegradation results for cellulose-positive
control. Standard error and 95% confidence interval values were calculated using
following inputs:

n1 = number of cellulose positive control samples = 2

n2 = number of compost control samples = 2

t = t-distribution value for 95% confidence interval with 2 degrees of

freedom = 4.303

239

Thus, standard error (8.) and 95% confidence interval of percent—
biodegradation of cellulose samples was 0.37% and 61.42% :I: 1.59%

respectively.

Table 5.6 Cumulative amount of carbon produced-Cellulose positive control

 

 

 

 

 

Cumulative carbon produced (grams)/
Sample ID
Percentage (%)
48 43.34 9
8A 43.59 9
Mean 1: Standard Deviation 43.46 :I: 0.18
Percent-biodeg radation 61 .42%

 

 

 

 

5.6.3.3 Results-Polypropylene-Glass negative control

Biodegradation results for PP-Glass composites (Polypropylene- (30wt%)
Glass fiber) and compost control (for comparison) are presented below in Figure
5.13. Referring to the 002 evolution curve, only sample 1B has C02 evolution
comparable to compost control sample for this study. Other two replicates of PP-
Glass negative control had lower cumulative C02 produced as compared to
compost control sample, even though negative control and compost control
samples contained equal amounts of compost subjected to similar composting
conditions. This result indicates PP-Glass composites are non-biodegradable
and can act as a hindrance to the existing composting environment. Cumulative
amount of carbon produced during the study and percent-biodegradation of PP-

Glass composites is presented below in Table 5.7.

240

Cumulative C02_PP-Glass-negative control

...13 H
60 at + 11A

 
 

 

+5A

 

 

 

5° “‘ -- compost blank

 

 

30*

 

 

 

Cumulative C02 (grams)
N
O

 

 

10

 

 

 

0 1O 20

 

1% r I I

30 Days 40 50 60

Figure 5.13 C02 evolution time plot for PP-Glass composites negative control

Table 5.7 Cumulative amount of carbon produced-PP-Glass negative control

 

Cumulative carbon produced (grams)/

 

 

 

 

 

 

 

Sample ID
Percentage (%)
18 17.54
11A 15.26
5A 12.75
Mean :1: Standard Deviation 15.18 :1: 2.40
Percent-biodegradation 0%

 

241

 

5.6.3.4 Results-PHB-Kenaf Composites

PHB— (30wt%) Kenaf composites were tested in triplicates and 002
evolution results for these samples and compost control are presented below in
Figure 5.14. As per the time plot below, their was variation in the cumulative
evolution of carbon dioxide from PHB-Kenaf composite samples. This variation
can be attributed to the processing technique for these composites; extrusion
and injection molding which produce a random mix of fibers with matrix. And,
since kenaf fibers with significant amount of lignin content are less biodegradable
as compare to PHB (Polyhydroxybutyrate) biopolymer, such random mixing of
kenaf fibers with PHB caused the variation in biodegradation results for PHB-

Kenaf composite replicates.

Cumulative C02_PHB-Kenaf Composite

 

 

 

 

 

 

 

 

 

 

 

160 _. . .. . .. - . _. _. . __-__ - __ __ _,_._-_._,_._ __.._.z_-_ ._._.._, , _. -....._____- j
+ 13A

140 ‘r + 9A /
E120 -- "T 3A A
g + compost blank W
3 100 - — 3
N I
O f/II/ 3
0 30 / l
e z
,3 M 7
E 60 .
E f, 3
3 40
o r

20 ‘

0 “ A r r r I r r

0 1o 20 30 Days 40 50 60

Figure 5.14 C02 evolution time plot for PHB-Kenaf composites

242

Cumulative carbon produced during the study for PHB-Kenaf composites

and percent-biodegradation of these composites is reported below in Table 5.8.

Table 5.8 Cumulative amount of carbon produced-PHB-Kenaf composites

 

 

 

 

 

 

Cumulative carbon produced (grams)l
Sample ID
Percentage (%)
13A 32.21
9A 41.97
3A 34.90
Mean :l: Standard Deviation 36.36 :I: 5.04
Percent-biodegradation 35.58%

 

 

 

 

Since, their was variation in the cumulative evolution of CO; from three
PHB-Kenaf composite samples, 90% confidence interval instead of 95% was
calculated in addition to standard error calculation using following inputs:

n1 = number of PHB-Kenaf composite samples = 3

n2 = number of compost control samples = 2

t = t-distribution value for 90% confidence interval with 3 degrees of
freedom = 2.353

The standard error (8.) and 90% confidence interval of percent-
biodegradation of PHB-Kenaf composite samples was calculated as 5.53% and

35.58% i 13.02% respectively.

243

5.6.3.5 Results-PHB-g MA-PHB-Kenaf Composites

In addition to PHB- (30 wt%) Kenaf composites tested for biodegradation,
PHB- (30 wt%)Kenaf composites with 5 wt% MA-PHB as compatibilizer were
also tested for biodegradability. The main objective behind testing compatibilized
PHB-Kenaf composites was to determine any variation in the biodegradation rate
of these composites as compared to PHB-Kenaf composites. These composites
were also tested in triplicates and the cumulative C02 evolution results for these

samples are presented below in Figure 5.15.

Cumulative C02_PHB-g MA-PHB-Kenaf Composite
160 ......FEH

+10A If

 

 

140 -_ 4‘ 6A Fi/
_,_ -— 4A R
-- compost blank M

.3
N
o

_L
o
o

 

03
O

 

 

Cumulative 002 (grams)
3 8

 

20‘

 

 

 

O 10 20 30 Days 40 50 60

Figure 5.15 002 evolution time plot for PHB-g MA-PHB-Kenaf composites

C02 evolution time plot above, indicates variation in the cumulative
amount of C02 produced for three PHB-g MA-PHB-Kenaf samples, and such

variation can be attributed to the processing technique, which was same as the

244

processing technique for PHB-Kenaf composites. However, this variation was
more than the variation in similar readings for PHB-Kenaf composites.
Cumulative amount of carbon produced and percent-biodegradation of these

samples is reported below in Table 5.9.

Table 5.9 Cumulative amount of carbon produced-PHB-g MA-PHB-Kenaf
composites

 

 

 

 

 

 

Cumulative carbon produced (grams)/
Sample ID
Percentage (%)
10A 28.37
6A 41.32
4A 41.40
Mean i Standard Deviation 37.03 :I: 7.50
Percent-biodegradation 37.35%

 

 

 

 

Standard error (8.) and 90% confidence interval of percent-biodegradation
of PHB-g MA-PHB-Kenaf samples was calculated as 8.33% and 37.35% :I:
19.61% respectively, using following inputs:

n1 = number of PHB-g MA-PHB-Kenaf composite samples = 3

n2 = number of compost control samples = 2

t = t—distribution value for 90% confidence interval with 3 degrees of

freedom = 2.353

245

5.6.3.5 Results-Neat Polymer samples

Biodegradation profile of three biopolymers: PHB (Polyhydroxybutyrate),
PLA (Poly-lactic acid), and CA- (30wt%) TEC (Cellulose acetate plaSticized with
Triethyl citrate) was obtained in this study. These polymers were selected for this
study because they are obtained from renewable raw materials and are some of
the currently studied biobased polymers. Additionally, it was useful to determine
the biodegradability of PHB polymer samples, in order to compare the results
with that of PHB-Kenaf composites. Such comparison was helpful in determining
the difference in biodegradability between PHB and Kenaf fibers.

These neat polymers were tested as single samples because of limited
available space in two glove boxes used for this study. Cumulative CO; time
plots for these samples are presented below, in Figure 5.16.

Based on the C02 evolution curves below, rate of biodegradation for neat
polymer samples (in descending order) was as follows: PHB > PLA > CA -
(30wt%)TEC. Cumulative carbon produced and percent-biodegradation for these

samples is reported below in Table 5.10.

Table 5.10 Cumulative amount of carbon produced-Neat Polymer samples

 

 

 

 

Sample Cumulative carbon Percent-
Polymer
ID produced (grams) biodegradation (%)
PHB 12A 45.00 51.15
PLA 7B 38.59 45.34
CA- (30wt%)TEC 10B 29.06 24.81

 

 

 

 

 

 

246

Cumulative C02_Neat Polymer samples

 

160 ~~ ... PHB

 
 
   

 

 

 

 

 

 

75140 —~ + PLA

g + CA-(30wt%)TEC
9 12° “" + compost blank
N

O

0

3

2.5

3

E

:r

0

 

 

 

0 1 0 20 30 40 50 60
Days

Figure 5.16 002 evolution time plot for Neat Polymer samples

At the end of the biodegradation study as recommended by ASTM D 5338
standard, pH of the compost-sample mix was determined. In addition dry solids
content, weight loss and visual observations of the compost-sample mix were
recorded. Finally, plant growth tests with cress seed and lettuce were conducted
in order to determine any adverse impacts of compost-sample mix on

germination of plants, once the final material at the end of the study is mixed with

soil.

247

5.7 End of study observatigns

To conclude the biodegradation study, a group of observations were
recorded including pH, dry solids content, weight loss, thickness, and pictures of
samples. pH of every compost-sample mix was recorded in order to confirm the
aerobic nature of the experiment. If a pH value of less than 7 is recorded, ASTM
D 5338 standard recommends measuring the volatile fatty acids content for the
compost-sample mix. And, in case volatile fatty acids content of more than 2 g
per kilogram of dry solids of compost-sample mix are recorded, that sample must
be regarded as invalid and rejected.

Composting vessels were weighed at the end of the study and dry solids
content of the compost-sample mix was determined. Thus weight loss of the
samples was calculated using final weight and dry solids content data. Weight-
loss data obtained was correlated to the cumulative C02 evolution, percent-
biodegradation data for the samples tested.

For a qualitative analysis of the extent of biodegradation, images of the
degraded samples were taken. Additionally, thickness of the samples (if possible)
was also recorded for a quantitative comparison between samples at the
beginning and end of the study.

Plant growth test for the final compost-sample mix was carried out as

recommended by ASTM D 6400 standard and based on OECD Guideline 208.

5.7.1 pH determination
Compost-sample mix pH was measured using the method developed by

Warncke20 (for additional information, please refer to Test methods for the

248

examination of composting and compost“). 500 mL plastic containers were ‘74
filled with compost-sample mix and distilled water was slowly added until the
sample was saturated. This saturated sample was allowed to equilibrate for 20
minutes and pH reading with standard laboratory pH meter was recorded. Since
the pH for the compost-sample mix was expected between 7 and 10, the pH
meter was calibrated with pH 7 and 10 standard solutions. pH readings recorded
are presented below in Table 5.11.

As noted in table 5.11, none of the samples had pH less than 7, which
indicates that none of the composting vessels went anaerobic. Therefore, no
volatile fatty acid measurements were recorded at the end of the biodegradation

study.

5.7.2 Dry solids and weight loss data

Total weight of the composting vessels was recorded at the beginning and
end of the composting study. Additionally, dry solids content of the samples at
the end of the study was used to calculate weight loss of samples during the
biodegradation study. Dry solids content and weight loss data is presented below
in Table 5.12. Correlations between weight loss and percent-biodegradation data

of test-samples are presented in the discussion section of this chapter.

249

Table 5.11 End of study pH readings

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Sample ID Sample Description pH Reading
7A Compost control 7.649
8B Compost control 7.587
8A Cellulose positive control 7.859
48 Cellulose positive control 7.795
5A PP-Glass negative control 7.725
11A PP-Glass negative control 7.747
1B PP-Glass negative control 7.692
3A PHB-Kenaf 7.652
9A PHB-Kenaf 7.774
13A PHB-Kenaf 7.637
4A PHB-g MA-PHB-Kenaf 7.768
6A PHB-g MA-PHB-Kenaf 7.830
10A PHB-g MA-PHB-Kenaf 7.756
12A PHB 7.914
7B PLA 7.394
108 CA- (30wt%)TEC 7.545

 

250

 

Table 5.12 Dry solids content and weight loss of samples at the end of

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

biodegradation study
Dry solids
Sample ID Sample Description Weight loss (9)
content (%)

7A Compost control 63.88 ND‘

8B Compost control 95.41 20.60
8A Cellulose positive control 61.25 80.93
48 Cellulose positive control 87.50 72.79
5A PP-Glass negative control 67.65 12.74
11A PP-Glass negative control 81.62 5.91

18 PP-Glass negative control 90.18 8.32

3A PHB-Kenaf 86.31 67.13
9A PHB-Kenaf 62.02 43.97
13A PH B-Kenaf 64.51 63.07
4A PHB-g MA-PHB-Kenaf 63.35 49.22
6A PHB-g MA-PHB-Kenaf 67.87 71.49
10A PHB-g MA-PHB-Kenaf 61.01 63.41
12A PHB 55.45 51.84
7B PLA 96.89 82.78
103 CA- (30wt%)TEC 97.33 41.72

 

 

 

 

 

 

‘ Not determined because of error in final weight reading

251

 

5.7.3 Sample images and thickness

Thickness of the dogbone shape tensile coupons of composites and neat-
polymers was measured at the beginning and end of the biodegradation study.
Since no biodegradation was observed for PP-Glass composite test samples,
their was no change in the thickness of these samples. For other samples, end of
study average thickness and percent-reduction data is presented below in Table

5.13.

Table 5.13 End of study sample thickness

 

 

 

 

 

 

Original Average end of study Average
sample sample thickness percent
Sample Description
thickness (mm) :1: standard reduction in
(mm) deviation thickness (%)

PHB-Kenaf 3.18 2.72 t 0.06 14.39
PHB-g MA-PHB-Kenaf 3.22 2.51 :l: 0.35 21.94
PHB 3.18 2.831017 11.11
PLA 3.25 2.72 :I: 0.11 16.23
CA- (30wt%)TEC 3.21 3.09 3.59

 

 

 

 

 

 

Images of compost-sample (PP-Glass negative control, dogbone shape
test samples) mix were taken at the end of the study, in order to qualitatively
analyze the extent of biodegradation. These images are presented below along—

with additional observations and explanation.

252

PP-Glass negative control
PP-Glass composite samples were used as a negative control for the

biodegradation study and as expected composite samples did not show any
signs of deterioration. Images of samples 5A and 11A are presented below in

Figures 5.17 and 5.18.

 

253

PHB-Kenaf composites
PHB-Kenaf composites showed significant degradation, which was

evident from significant microbial growth and deterioration of composite samples.
Though composite samples had weakened, degradation was still surface limited
indicating that the significant degradation could be carried out for more than two
months. Degradation images of sample 3A, 9A, and 13A are presented below in

Figures 5.19, 5.20, and 5.21 respectively.

 

Figure 5.19 Sample 3A PHB-Kenaf composites-End of study image

254

 

Figure 5.21 Sample 13A PHB-Kenaf composites-End of study image

255

PHB-g MA-PHB-Kenaf composites
PHB-g MA-PHB-Kenaf composites had similar levels of degradation as

compared to PHB-Kenaf composites at the end of biodegradation study. End of
study images for samples 4A, 6A, and 10A are presented below in Figures 5.22,
5.23 and 5.24 respectively. Deterioration and microbial growth on the samples

was evident as can be seen in the images below.

 

   

..z ‘
~_ 0 t

Figure 5.23 Sample‘6A PHB-g MA-PHB-Kenaf composites-End of study image

256

 

Figure 5.24 Sample 10A PHB-g MA-PHB-Kenaf composites-End of study image

Neat polymer samples
Neat polymer samples exhibited varying levels of degradation by the end

of biodegradation study and such differences were apparent from the images
taken at the end of study. PHB samples had significant microbial growth and
were disintegrating into smaller segments; however degradation process could
be carried out for more than two months. For PLA samples even though
microbial growth was not evident, samples had become brittle and were
disintegrating into smaller segments. Surface degradation was present for CA-
(30wt%) TEC samples; however there was no disintegration and no significant
reduction in strength of these samples. End of study images for PHB, PLA, and

CA- (30wt%) TEC are presented below in Figures 5.25, 5.26, and 5.27.

257

 

Figure 5.26 Sample 7B PLA-End of study image

258

 

Figure 5.27 Sample 10B CA- (30wt%)TEC-End of study image

5.7.4 Plant growth test

OECD Guideline 208 “Terrestrial Plants, Growth Test”22 was followed to
determine the toxic effects of materials tested in this study on early stages of
growth of various terrestrial plants. OECD guideline recommends testing
samples in varying concentrations mixed with soil, with minimum of four
replicates per sample, and minimum of five seeds per replicate. A minimum of
three plant species should be tested with at least one from each of the three
categories defined by OECD Guidelines represented in terms of percent

emergence of plants and of growth in plant-weight.

259

In the current study, compost-sample mixes were directly tested, without
mixing with soil. The samples were tested with four replicates, with each replicate
comprising of 25 cells in a 200-cell tray and a total of 25 seeds per replicate.
Each cell contained approximately 7 g of compost-sample mix. These samples
were watered daily and at the end of study total number of cells in which seeds
germinated was recorded, in order to calculate percent germination.

Initially plant growth test was started with cress seeds; however because
of concerns over cress-seed quality, additional tests with lettuce seeds were
conducted. It should be noted that cress and lettuce seeds belong to category 3
of plant species defined by OECD guidelines and no seeds were selected from
other two categories because of insufficient amount of available compost-sample
mix. Additionally plant growth tests could not be conducted for PLA and CA-
(30wt%) TEC samples, because of insufficient amount of sample available at the
end of biodegradation study.

Percent germination for cress seed samples was recorded after 3 1/2
weeks, while for lettuce seed samples it was recorded after 2 1/2 weeks. Percent
germination results for cress seed and lettuce seed are presented below in
Tables 5.14 and 5.15 respectively.

As per the germination results below, PHB-Kenaf, PHB-g MA-PHB-Kenaf,
and PHB samples had higher percent germination as compared to PP-Glass
negative control samples and slightly higher or comparable germination
compared to Compost control samples; thus indicating the suitability of PHB

polymer and its natural fiber reinforced composites to support plant growth at the

260

end of biodegradation study. Cellulose positive control samples had the lowest
percent germination among the samples tested. This unexpected result could be
attributed to the small particle size of cellulose (13 pm), which acted as a binding
agent causing the compost-cellulose mix to clump when water was added,

therefore hindering the plant germination.

Table 5.14 Cress seed-percent germination results

 

 

 

 

 

 

 

 

 

Cress seed (25 seeds per sample replicate)
S , , Average percent
ample description Replicate Replicate Replicate Replicate
germination :I:
1 2 3 4
standard deviation
Control‘ 92 88 96 88 91 :I: 4
Compost control 84 72 80 68 76 :I: 7
Cellulose positive
68 72 68 64 68 :I: 3
control
PP-Glass negative
68 76 84 68 74 i 8
control
PHB-Kenaf 84 76 80 84 81 :I: 4
PHB-g MA-PHB-
76 76 80 76 77 :I: 2
Kenaf
PHB 72 80 84 68 76 :I: 7

 

 

 

 

 

 

 

 

 

‘ Scotts® potting mix was used as a control for this test

261

Table 5.15 Lettuce seed-percent germination results

 

 

 

 

 

 

 

 

 

Lettuce seed (25 seeds per sample replicate)
Sample Average percent
. I' l' R l' . .
description Replicate Rep Icate Rep rcate ep rcate germination :I:
1 2 3 4 standard
deviation
Control‘ 96 100 100 92 97 t 4
Compost
96 92 96 88 93 :I: 4
control
Cellulose
positive 84 88 84 76 83 :I: 5
control
PP-Glass
negative 88 84 84 88 86 :t 2
control
PHB-Kenaf 92 96 96 96 95 :I: 2
PHB-g MA-
88 100 96 92 94 :I: 5
PHB-Kenaf
PHB 84 96 100 92 93 :I: 7

 

 

 

 

 

 

 

 

 

' Scotts® potting mix was used as a control for this test

262

5.8 Qscpssion

The motivation behind this study was to understand the behavior of
biobased composites and polymers under composting environment. These
results are useful since they can be used to establish the superior end of life
profile of biobased composites over conventional composites and C02-neutral
nature of biobased composites.

The cumulative carbon numbers for compost control and cellulose positive
control replicates had very less standard deviation; thus indicating good quality
control, which was evident from a smaller 95% confidence interval for percent
biodegradation of cellulose samples. ASTM D 5338 standard (based on which
this study was conducted) recommends a minimum of 70% degradation of
cellulose positive control samples after 45 days. However, in the current study
61% average degradation was achieved for cellulose samples in 60 days. This
reduction in degradation numbers can be attributed to two factors. First, the
compost used in this study was a well-matured compost (see section 5.4) with
minimal 002 emissions and thus this compost might have retarded the normal
rate of biodegradation as recommended by ASTM D 5338 standard. Second,
cellulose10 used for the study had a particle size of 13 um and such small particle
size acted as a binding agent and lead to agglomeration of compost-cellulose
mix, thus reducing the porosity of compost-cellulose mix and hindering the
degradation of cellulose positive control samples.

To improve the porosity of compost-sample mix, Perlite19 was added;

however, the results obtained were mixed as indicated by low percent-

263

biodegradation of positive control samples. For future studies, other porosity-
enhancing inert materials such as wood chips, gravel could be used.

PP-Glass composite samples (100 grams PP-Glass composites + 600
grams compost control, dry solids only) were tested as negative control in this
study and average cumulative C02 production for these samples was lower than
002 production for compost control samples (600 grams of compost control, dry
solids only). This result indicates that if non-degradable component (such as PP-
Glass) is mixed with compost control, it can reduce the original CO; production
from compost and therefore hindering the biodegradation process.

PHB-Kenaf and PHB—g MA-PHB-Kenaf composite samples were still
actively degrading at the end of study and exhibited variation among replicates
for degradation results. Such variation can be attributed to the processing
(extrusion and injecting molding) of these composites, which induces a random
mixing of kenaf fibers (less degradable) with PHB biopolymer (readily
degradable).

Percent-biodegradation of selected neat biopolymer was calculated, with
rate of biodegradation (in descending order) as follows: PHB > PLA > CA- (30
wt%) TEC. PHB samples had a percent-biodegradation of 51% and using this
number, percent—biodegradation of kenaf fibers in PHB-Kenaf composites was
calculated‘. This calculation indicates that kenaf fiber portion of composites is

essentially non-degradable, which is due to high Iignin content of kenaf fibers”.

 

' 100 grams of PHB-Kenaf composites have 70 grams of PHB. 70 grams of PHB have 37.87
grams of carbon (54.10% carbon content). Using 51.15% biodegradation rate for PHB, 19.37
grams of carbon is produced. Average carbon produced for PHB-Kenaf composites is 36.36

grams and with average carbon production of 17.63 grams from compost, 18.73 grams is the

264

The extent of biodegradation of the samples was determined based on
two parameters: percent biodegradation and weight loss. Correlation between
these two parameters is presented in Table 5.16. Ideally, percent biodegradation
of a sample should have a positive correlation with weight loss; but in the current
study this is not the case. Additionally, cumulative carbon produced and weight
loss recorded have different values, when ideally it should be the same number.

These deviations can be attributed to two factors: 1) biodegradation
mechanism of test-samples 2) gas sampling procedure and frequency.

During biodegradation, complex polymeric samples are broken down into
oligomers, and generally the end products are C02 and H20. For example
degradation of PHB (as described by Albertsson and Karlsson“, shown below in
figure 5.28) takes place through TCA (tricarboxylic acid cycle) cycle, producing
C02 and H2025. Since the gas chromatograph readings of the test-samples only
measure 002, water produced during biodegradation partially accounts for the
difference between cumulative carbon produced and weight loss of test-samples.

Gas sampling of the test-samples was done approximately once a day
and point reading of C02 percentage in combination with exhaust gas flow rate
was converted to cumulative C02 produced/day, as explained in Appendix 5.6.3.
This calculation assumes that the rate of C02 production remained constant
during that time-period. Therefore, this assumption might also account for the

difference between cumulative carbon produced and weight loss of test-samples.

 

contribution from PHB-Kenaf composites. As we can see theoretically calculated contribution
from PHB approximately matches the experimental carbon contribution from PHB-Kenaf
composites!

265

Despite these deviations in the biodegradation study, the final results do
not conflict with the objectives of the study, which were to determine the rate of
biodegradation and environmental impacts of degradation end-products. This
study should be considered preliminary, and is currently being repeated under
similar controlled composting conditions at Composite Materials and Structures

Center (CMSC) at Michigan State University.

 

 

 

 

 

 

CH3COOH Glucose Fatty acids CH30H2CH2COOH
Fructose
CoASI-I ATP
ATP + CoASI-l
> AMP + PPI
AMP + PPI f -O CH3C Hzc Hch-SCoA
FAD

v y non,
mm CH3CONSCoA CH30H=CHCO-SC0A

I; H20

NAD“

qI-r

CH3COCH2CO$C0A M; (L)CH;CHCH200-SCOA
NADPH ®®
AMP
NADP+ G) PPI
(D)CH3CHCI-IzCO-SC0A ATP cH3cocrr2coon
COASH ¢
H
./ (9 NADH<-\
CoASI-I i N AD’ / (D
t l
P(3HB) F (D)CH3CHCFhCOOH
I
OH

Figure 5.28 Cyclic metabolic pathway of the biosynthesis and degradation of
P(3HB), reproduced from Albertsson, Karlsson (1995), see ref. 24

266

In the current study, dogbone samples of composites and polymers were
reduced to an approximate size of 2 cm x 2 cm x 3.2 mm and the rate of
biodegradation could have been improved by further reducing the sample size,
thus increasing the surface area of samples exposed to microbial degradation.
Such improvements in the rate of biodegradation have been previously recorded
by Buchanan et al.26, where polymer films (0.051 to 0.203 mm thick) have been
tested under controlled composting environment. For example: PHB films had
75.8% weight loss in 14 days, and 70/30 CAfl' EC films had 100% weight loss in
12 days, while 70/30 CA/T EC injection molded bars (3.1 mm thick) lost 10.1% of
their weight under controlled composting cycle of 30 days.

Addition of surfactants to reduce hydrophobicity and increase surface area
has been used to increase the rate of biodegradation of powdered polyethylene

L“, addition of

in biotic samples”. However, in another study by Walter et a
surfactants to poly(trimethylene succinate) film samples reduces the rate of
biodegradation. Such a behavior is due to the fixed surface area of solid film
samples, and adding surfactants only reduces the hydrophobicity of film samples
and hydrophobic character of samples is known to aid enzymatic degradation”.

Therefore, addition of surfactants may increase rate of biodegradation, only if the

test samples are in powder form.

267

Table 5.16 Correlation between cumulative carbon produced/percent
biodegradation and weight loss at the end of the study

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Cumulative
Sample Percent- Weight
Sample Description carbon
ID biodegradation loss (9)
produced (9)

7A Compost control 17.53 NAii ND'

8B Compost control 17.72 NA" 20.60
8A Cellulose positive control 43.59 61.71 80.93
4B Cellulose positive control 43.34 61.12 72.79
5A PP-Glass negative control 12.75 0.00 12.74
11A PP-Glass negative control 15.26 0.00 5.91

13 PP-Glass negative control 17.54 0.00 8.32

3A PHB-Kenaf 34.90 32.81 67.13
9A PHB-Kenaf 41.97 46.24 43.97
13A PHB-Kenaf 32.21 27.70 63.07
4A PHB-g MA-PHB-Kenaf 41.40 45.76 49.22
6A PHB-g MA-PHB-Kenaf 41.32 45.60 71.49
10A PHB-g MA—PHB-Kenaf 28.37 20.69 63.41
12A PHB 45.00 51.15 51.84
78 PLA 38.59 45.34 82.78
10B CA- (30wt%)TEC 29.06 24.81 41.72

 

 

“ Not Applicable
‘ Not determined because of error in final weight reading

268

 

References:

 

‘ A. Keller et al., Degradation kinetics of biodegradable fiber composites. Journal of
Polymers and the Environment, Vol. 8, No. 2, 2000.

2 Rynk, Robert, editor. On-Farm Composting Handbook. Ithaca, New York: Northeast
Regional Agricultural Engineering Service. 1992.

3 R. Narayan, Drivers for biodegradable/compostable plastics 8 role of composting in
waste management & sustainable agriculture. Bioprocessing of Solid Waste & Sludge,
Vol. 1, 2001.

4 A. Stamecker, M. Menner, Assessment of biodegradability of plastics under simulated
composting conditions in a laboratory test system. International Biodeterioration &
Biodegradation (1996) 85-92

5 Gattin, R. et al., Comparison of mineralization of starch in liquid, inert solid and
compost media according to ASTM and CEN norms for the composting of packaging
materials. Biotechnology letters 22: 1471-1475, 2000.

6 Copinet, A. et al., Photodegradation and biodegradation study of a starch and
poly(lactic acid) coextruded material. Journal of Polymers and the Environment, 11(4),
2003.

7 R Jayasekara et al. (2001), Journal of Chemical Technology and Biotechnology

3 Test methods for the examination of composting and compost (T MECC), Section
04.02-D, Total Nitrogen by Oxidation. May 11, 2002.

9 Official Solvita® Guideline-Compost Maturity Index, Version 4.0, Woods End Research
Laboratory, Inc. Mt Vernon ME. URL: http:llwww.solvita.com/aaa/solvita.html

1° Lattice® Microcrystalline Cellulose, Type: NT-013, FMC BioPolymer, Philadelphia, PA,
USA.

‘1 Pro-fax 6523, Polypropylene homopolymer, Basell North America Inc., Elkton, MD,
USA.

‘2 Chopped Strand 756F, V4” length, Johns Manville, Waterville, OH, USA.

‘3 Biomer® P226, Poly((R)3-hydroxybutyric acid) pellets, Biomer, Krailling, Germany.

‘4 Kenaf fiber (un-chopped), Flaxcraft, NJ, USA.

‘5 Biomer® L9000, Poly-L-Iactate pellets, Biomer, Krailling, Germany.

1° Powdered Cellulose Acetate (CA), plasticizer and additive free, degree of substitution
of cellulose acetate was approximately 2.5. Eastman Chemical Company, Kingsport, TN,
USA.

‘7 Triethyl Citrate (TEC), Morflex Inc., Greensboro, NC, USA.

‘8 Standard Methods for the examination of water and wastewater, 19th Edition, 1995,
American Public Health Association, Washington, DC.

‘9 Perlite is a natural volcanic glass; it can be classified as amorphous mineral containing
fused sodium-potassium-aluminum silicate. Definition obtained from www.cdc.gov.
Additional information available at www.cerliteorg

2° Wamcke, D. 1998. Greenhouse root media. pp. 61-64. In Recommended chemical
soil test procedures for the North Central Region. North Central Regional Research
Publication No. 221 (Revised) Missouri Agricultural Experiment Station SB 1001.

2‘ Test methods for the examination of composting and compost (T MECC), Section
04.11 Electrometric pH determinations for compost. March 21, 2002.

22 OECD Guideline for testing of chemicals, adopted April 1984. Available from
Organization for Economic Development (OECD), Director of Information, 2 rue Andre’
Pascal, 75775 Paris Cedex 16, France.

269

 

23 Mohanty, A.K. et al., Biofibres, biodegradable polymers and biocomposites: An
overview. Macromolecular Materials and Engineering, 276/277, 1-24 (2000)

2‘ A.-C. Albertsson and S. Karlsson, Degradable polymers for the future. Acta Polymer.,
46, 114-123 (1995)

25 H.-M. Muller and D. Seebach, Poly(hydroxyalkanoates) : A Fifth Class of
Physiologically Important Organic Biopolymers? Angew. Chem. Int. Ed. Engl. 1993, 32,
477-502

26 Buchanan, C.M. et al., Biodegradation of Cellulose Esters: Composting of cellulose
ester-diluent mixtures. Journal of Macromolecular Science — Pure and Applied
Chemistry, A32(4), 683-697 (1995)

27 Karlsson, S., Ljungquist, 0. and Albertsson, A.-C. Biodegradation of polyethylene and
the influence of surfactants. Polym. Degr. Stab.1988, 21, 237-250

28 Walter, T. et al., Enzymatic degradation of a model polyester by lipase from Rhizopus
delemar. Enzyme and Microbial Technology 17:216-224. 1995

29 Blow, D. Lipases reach the surface. Nature 1991, 351, 444-445

270

Chapter 6: Conclusions and Future Work
6.1 _L_CA Sm; conclpsions

A cradle to pellet LCA study for production of 1000 tensile coupons of
PHB-(30wt%)Kenaf and PP-(30wt%)GIass composites was conducted in the
present research work. Thus, effectively making a volumetric comparison for both
the composites. Resource consumption and emissions for both the systems were
computed as part of the LCI (Life Cycle Inventory) analysis and for a better
understanding of this data, CML impact assessment methods were used to
classify inventory flows based on their environmental impacts.

In 7 out of 9 impact categories evaluated, PHB-Kenaf composites had
higher impacts compared to PP-Glass composites, with lower impacts for global
warming and abiotic resource depletion categories. Normalizing the impact
assessment results based on the factors defined for the World in 1995, four
impact categories of Abiotic resource depletion, Acidification, Eutrophication, and
Global warming accounted for more than 90% of the cumulative normalized
impacts. Normalization made it possible to have a quantitative comparison of
various impacts, whereas a qualitative comparison of the impacts can be done
using weighting or valuation methods. However, because of the unavailability of
widely acceptable valuation/weighting methods, only quantitative comparison of
the impacts was done for the current study.

Nutrient outflows (N&P) from soybean cultivation are important
contributors towards eutrophication and acidification impacts. Since corn and

soybean are rotational crops, nutrient outflows from them are related. In the

271

default allocation approach (C-S allocation), approximately equal N-application
and leaching rates were considered for corn and soybean crops; while in the
alternative C-allocation approach, all N-application and emissions were allocated
solely to corn cultivation. As determined in the sensitivity analysis, using C-
allocation approach, all impacts were lower compared to C-S allocation
approach, with significant reduction of 47% for eutrophication impacts. Therefore,
cumulative normalized for PHB-Kenaf composites range between +3% (C-S
allocation) to -9% (C-allocation) of the cumulative normalized impacts for PP-
Glass composites.

In addition to analyzing the sensitivity of LCA results due to different
allocation methods, effect of different eutrophication impact assessment method
on final results was also analyzed. Contrary to global impacts such as climate
change, eutrophication is considered to be a regional and/or local impact. The
CML eutrophication method computes the impacts without considering location
related variation, where such a variation can influence the actual impacts due to
site-specific emissions. An alternative method developed by EPA, TRACI
considers such variation by multiplying CML eutrophication potentials with US
state-level nutrient transport factors, and computing US weighted-average
eutrophication potentials, and in comparison to CML eutrophication potentials
were significantly lower for air emissions, but remained the same for water
emissions of nutrients. Thus, TRACI eutrophication impacts for PHB-Kenaf
composites using C-S and C-allocation methods were lower by 38% and 68%

respectively compared to impacts calculated using CML eutrophication method.

272

This reduction had significant effect on cumulative normalized impacts of PHB-
Kenaf composites, which ranged from -2% (C-S allocation) to -14% (C allocation)
of those computed for PP-Glass composites.

Comparing energy consumption, PHB-Kenaf composites used 22% (for C-
S allocation) and 25% ( for C-allocation) lower energy compared to PP-Glass
composites, mainly due to low energy consumption for kenaf fiber production
(97.6% reduction compared to glass fiber production) and high energy
consumption for PP production process.

Summarizing LCA results, majority of environmental impacts and energy
consumption for PHB-Kenaf composites were due to PHB production, specifically
from electricity and steam production, and nutrient emissions from soybean
cultivation. Unlike PP production, technology for PHB production is in its infancy
and increased research emphasis and commercial production will definitely

improve its efficiency, thus reducing material and energy requirements.

6.2 Biodegradation sttflv conclpsions

The motivation behind biodegradation study was to determine the rate of
biodegradation of biobased composites and polymers under controlled
composting environment, and study the environmental impacts of degradation
end-products. These results are useful since they can be used to establish the
superior end of life profile of biobased composites over conventional composites
and 002-neutral nature of biobased composites.

PP-Glass composites were tested as negative control (100 grams PP-

Glass composites + 600 grams compost control, dry solids only) in this study,

273

and cumulative COz production for these samples was lower as compared to
compost control samples (600 grams of compost control, dry solids only),
indicating a hindering effect of negative control sample on the biodegradation
process. At the end of the study in 60 days, PHB-Kenaf composites had
degraded by 36% and still actively degrading, where the degradation might have
reached appreciable levels (70% or above) in six months, if the experiment was
continued. Among the neat polymer samples, PHB had the highest percent-
biodegradatlon of 51 %, followed by 45% for PLA and 25% for CA-(30wt%)TEC.

Environmental impacts of degradation end-products were determined by
recording percent germination of cress and lettuce seed in compost-sample
mixes (using OECD Guideline 208). Based on the germination results, PHB-
Kenaf, PHB-g MA-PHB-Kenaf, and PHB samples had higher germination
compared to PP-Glass samples and comparable or slightly higher germination
than compost control samples; thus Indicating the suitability of PHB polymer and
its natural fiber reinforced composites to support plant growth at the end of
biodegradation study.

Finally, in order to improve the rate of biodegradation, using an active
compost and further reduction of the sample size (in the current study, it was

approximately 2 cm x 2 cm x 3.2 mm) is recommended.

6.3 Future work and Recommendations
LCA and biodegradation evaluation for the current study were conducted
for composites with equal fiber weight fraction (30wt%) and comparable tensile

modulus. And, composites with higher natural fiber weight fraction were not

274

processed because of the limitations due to automatic fiber addition during
extrusion processing of composites. Future studies should develop
biocomposites with higher fiber content and evaluate the impacts of varying fiber
weight fraction on mechanical properties of biocomposites. The past studies by
Sanadi et al.1, Nishino et al.2, and Wambua et al.3, clearly shows a direct
relationship between increase in natural fiber weight fraction and mechanical
properties of the resulting natural fiber composites. As an example, Figure 6.1
(Wambua et al.3) shows an increase in tensile modulus of PP-Kenaf composites

with increasing kenaf fiber weight fraction.

10
9.
3.

7.1 I
6. {/t/
5.

4.
3.
2.‘
1.

01*V'TTTTV'11
o 10 2o 30 40 so 60

Fiber weight fraction (wt%)

 

Tensile Modulus (GPa)

 

 

 

Figure 6.1 Effect of fiber weight fraction on tensile modulus of PP—Kenaf fiber
composites (Wambua et al.3)

Additionally, future studies should also evaluate the effect of increased
fiber loading on sustainability of these composites. Since, the production of kenaf
fibers has significantly lower energy consumption and environmental impacts

compared to glass fibers, higher fiber reinforcement will lead to a significantly

275

weight (on volume basis) compared to PP-(30wt%)Glass composites. However,
as shown below in Figure 6.2 (based on a theoretical calculation), with fiber
loading of 45 wt% or more, PHB-Kenaf composites have weight advantage over
PP-Glass composites. This weight advantage can have a significant effect on the
LCA results, if the use-phase of composites is included in the study. For the
current study use-phase was excluded because of unavailability of production
data for composites used as automotive structural components. Therefore, for
future LCA studies of biobased composites, partnership with automotive
companies should be considered, so as to obtain accurate product information

and study multiple scenarios with varying fiber composition.

 

 

12 ~
I’-
’a
’I
11 -I . eeeeeeeeeeeeeeeeeee ‘eooe’e'e’ eeeeeeeeeee .
F. .3”
M ,’
:5. ’,I’
H -1 ”
f” 10 l” ....... PHB-Kenaf
0
3 ---- PP-Glass
9 .
8 l T I
30% 40% 50%

Fiber weight fraction (wt%)

Figure 6. 2 Weight difference (volume basis) for PHB-Kenaf and PP-Glass
composites

276

Even though disposal phase was not included in the current study, it was
indirectly evaluated by testing the biodegradability of composites under controlled
composting environment. The composting study clearly shows a superior end of
life profile for PHB-Kenaf composites. The future LCA studies of composites
should consider multiple disposal scenarios such as Iandfilling, incineration, and
recycling in addition to composting. However, the biodegradable nature of PHB-
Kenaf composites makes composting the preferred disposal option; while
incineration and recycling are suitable for non—degradable PP-Glass composites.

Finally, future LCA studies should focus on obtaining more consistent and
updated data. The inventory analysis for the current LCA study used data from a
wide variety of sources such as DEAMT'V' database, APME, and journal
publications and even though modifications were done to make it relevant to US
production, using a single database for inventory analysis is the best option. The
recent availability of US life cycle inventory data from NREL4 is an encouraging

development in this direction.

277

References:

 

1 Sanadi AR, Caulfield DF, Jacobson RE, Rowell RM. 1995. Renewable Agricultural
Fibers as Reinforcing Fillers in Plastics - Mechanical-Properties of Kenaf Fiber-
Polypropylene Composites. Industrial & Engineering Chemistry Research 34(5):1889-
1896.

2 Nishino T, Hirao K, Kotera M, Nakamae K, lnagaki H. 2003. Kenaf reinforced
biodegradable composite. Composites Science and Technology 63(9):1281-1286.

3 Wambua P, Ivens J, Verpoest I. 2003. Natural fibres: can they replace glass in fibre
reinforced plastics? Composites Science and Technology 63(9):1259-1264.

‘ NREL. 2006. US. Life-Cycle Inventory (LCI) Database, http:llwww.nrel.gov/lcil.

278

2:802:40 oFmFm Beaming: £38200 .n_ 0505 ....2 Eat cocfifio .83 Lac.» .2 F33. 052605 .8 8950 F

.m004HQ wanna m.umw> umma co comma mum mwumefiumw Sofiuodfioua mo umoo

"muoz

 

Hm.mmm 9m00 DmHuHUmmm 44909
00.0 m0m<4 0m9<0044<ZD
v4.04 4<9Hm<0 UZH9<mmmo 20 9mmmm9ZH
04 mmm 00 bma 0N.va 009.4 mm.vm 0H ma mh.Hm MN.VH m44909
0N.vH vm.m 0vv.0 mv.m mo H 00.m mv.m has 00.H 0NN.0 an 004 .Nm .Nm umnafldm 043002
04.NH mm.4 0NN.0 Ho.m mm.0 00.m mv.m HMS 00.4 0NN.0 Q: 004 mawfl v >003m 440m
mm.0m mm.v 000.0 ma NH 0N m 0m.0 0N.v Mmz 00.4 000.0 0: 004 mum mm kumw>umm mmmHHm
00.0w 00.0w 00.0w 0000.H whom mwflfluflnuwm
00.0 00.0 00.0 0000.H 0:4 00.H 400m 00.0 Had >4 and
00.mH 00.mH 00.mH 0000.H whom MwNHHHuuwm
00 0 00 0 00.0 0000.H 034 00.4 0900 00.0 Hflfl >4 uuwm 004
00.0w 00.0w 00.0w 0000.4 430 00.4 whom coflummHHuH
mH.m mm.0 000.0 mm.0 0N.0 0H.H «0.0 C50 00.4 v00.0 Q: 004 0¢tmm wum>HuHSU
Nb.v 00.0 mm0.0 N0.0 0m.0 vw.H 90.4 >mz 00.4 mmo.0 an 004 ovumm wum>HuHSU
00.0w 00.0w 00.0w 0000.4 whom wam
mm.m 00.4 004.0 00.H m0.0 HN.H 00.0 >02 00.4 000.0 an 004 ovtmm ucmam
00.mN 00.mm 00.mN 0000.4 whum uwuflafiuuwm
00.v 00.v 00.v 0000.H >62 00.H whom 00.v uhwm >HQQ¢
00.mH 00.mH 00.mH 0000.H whom mwvfioflnuwm
00.v 00.v 00.v 0000.H >m£ 00.4 whom 00.0 Em£0 >HQQ<
m4.m vv.0 mm0.0 09.0 hm.0 40.0 00.0 Had 00.4 mm0.0 an 004 .bm .PNABOMMMIV UCOU 30m
mm.m 0N.H 0mH.0 0m.N 04.H mm.N 00.4 HQ< 00.N 090.0 an 004 .vm .vm BOHHMI Xmfio
llllll wHMHHOUIIIIIII mHQHHOU IIIIIIIIIIIWHGHHOUIIIIIIIIIIII

9m00 9m00 mUHmm 9ZDOE< 9m00 mmDOI meHm 90mmH0 meHm 90mmHQ :92 mm>0 m9<m mNHm 9HZD 9DmZH UZH9<mmmO
44909 ttttttttttttttttttttttttttttttttttttttttttttttttttttttttttttttt mm2H9 mmmm mo90¢m9 \mNHm \ZOH9dmmmo

9szH OZH9<mm00 m0m<4 00444 9000 QHDOm 9m00 m090<m9

 

.002 09.3. meow ..oa £60980 2o: ace. mFmoo 0cm cm: 8582 noFmEsmm ”adv 50:83.

279

Appendix 4.3.6: Energy requirement for Extrusion and Injection
molding

Emger enepcy regpirement

Electricity consumption for extruder drive and heater is the only form of
energy requirement for extrusion, and was categorized either as fixed energy
requirement (FER) or variable energy requirement (VER). FER accounts for the
extrusion steps, which remains same irrespective of the amount of composites
being processed, while VER accounts for the variation in the weight of the same.
The ZSK-30 twin-screw extruder was used for processing composites in the
current researchand the formula for calculating extruder heater and drive energy
requirements are mentioned below.

Actual dn've power requirement was calculated using this formula:

_ (RPM x Torque x 180)
9550 x 94%

kW

 

where RPM = Screw speed = 100 rpm (for the current study)

Torque = 40% (in 40-60% range, lower value selected)

Drive efficiency1 = 94% (1-2% motor and 4-6% gearbox losses)
Thus, drive energy requirement, ED = kW x t, where t is the time in
seconds to extrude composites.
Heater energy requirements2 were calculated based on the

thermodynamic properties of the materials being extruded as follows:

m X ((Cp X AT) + AHfusion)
75%

 

where m = mass of composites extruded in kilograms

280

Cp = specific heat of the material in KJ/kg.K
AT = Temperature rise during extrusion in Kelvins
AHfusion = heat of fusion in KJ/kg
Heater efficiency3 = 75% (accounting for losses to surroundings
and energy removed by cooling water)
Thermodynamic properties of the extruded materials are mentioned below

in Table Appendix 4.3.6.1.

Table Appendix 4.3.6.1 Thermodynamic properties of the extruded materials

 

 

Material Cp, KJ/kg.K AHmsion, KJ/kg
LDPE 2.30 218.60
PHB 3.18 101.50
Kenaffiberj 1.34 NA

PP 2.09 102.00
Glass fiber" 0.66 NA

 

 

' Specific heat of Kenaf fibers assumed to be same as Cellulose
" Specific heat of Glass fibers assumed to be same as Glass wool

281

Following assumptions were made for extruder energy calculations:
- Initial purge time of 5 minutes with the material being processed, so
as to avoid contamination with the previous material processed
- Final purge time of 5 minutes with the purge material (LDPE used
as purge material)
- Energy consumption for warm-up of the extruder not considered
- Extrusion feed rate of 57.10 grams/min
Based on these assumptions, fixed energy and variable energy

requirements were calculated as follows:
FER = (ED+ Q)lnitial Purge + (50+ Q)Final Purge
VER 2 (ED"' Q)Material
EExtmsion = FER + VER

Energy requirements for PHB-(30wt%)Kenaf and PP-(30wt%)Glass

composites are mentioned below in Table Appendix 4.3.6.2.

Table Appendix 4.3.6.2 Extrusion energy requirement (MJ-Electricity)

 

 

Amount of composites FER VER ngmion
Material AT (K)
extruded (kg) (MJ) (MJ) (MJ)
PHB-(30wt%)Kenaf 10.805 160 1.05 16.20 17.25
PP-(30wt%)Glass 9.805 155 0.99 12.57 13.55

 

282

Injection moldimnegv requirement

Energy requirement for injection molding is from two processes: shot
preparation and forming. As per Rosato et al.“, energy requirement for shot
preparation is generally an order of magnitude greater than required for forming.
The details for calculating these requirements are mentioned below.

Shot preparation step involves phase change of polymer from solid to melt
phase and is achieved by using barrel heaters. Energy requirement for this step
is calculated as follows:

ES = mx((Cp xAT)+AH,,,S,O,,), .

where m = amount of composites injection molded in kilograms
AT = Temperature rise during injection molding in Kelvins,
And, thermodynamic properties specified in Table 1 were used to
calculate energy requirement for shot preparation step.
Forming step requires drive mechanism, so as to mold the melt into final
product shape. Forming energy requirement is calculated as follows:

E _ Nx(PxQ)xfill time
F— 116379 ’

 

where Q is volumetric injection rate in cm3/sec, and calculated as:

Q: fill speed x Screw cross-sectional area,

Screw diameter = 3.20 cm,
P = injection pressure in psi = 1000 psi,
N = Number of tensile coupons molded = 1000

Shot size

Fill time = _
Frll speed

283

Thus, total energy requirement for injection molding is defined as follows:

: (ES + EF)
"‘4 25%

where Process efficiency = 25% (Rosato et al.4, have specified injection
molding efficiency in the range of 10-25%, higher value was selected for the

current study). The total energy values calculated for composites are mentioned

below in Table Appendix 4.3.6.3.

Table Appendix 4.3.6.3 Injection molding energy requirement (MJ-Electricity)

 

Amount of
Shot size Fill speed Es EF Em
Material composites AT (K)
(cm) (cm/sec) (MJ) (MJ) (MJ)
molded (kg)

 

PHB-(30wt%)Kenaf 10.805 140.56 0.95 0.25 4.76 0.17 19.72

PP-(30wt%)Glass 9.805 190.56 1.0 0.5 3.81 0.18 15.93

 

284

References:

 

1 Data obtained from Coperion Corporation (ZSK-30 manufacturer).

2 Extrusion: the definitive processing guide and handbook / by Harold F. Giles, Jr., John
R. Wagner, Jr., Eldridge M. Mount, Ill.

3 Process efficiency for extruders range from 35% to 75%, higher value selected. Ref:
Injection molding handbook / Dominick V. Rosato, Donald V. Rosato, Marlene G.
Rosato. 3rd ed.

4 Injection molding handbook / Dominick V. Rosato, Donald V. Rosato, Marlene G.
Rosato. 3rd ed.

285

Appendix 5.2.1: Wattage requirement for controlled composting
environment

Wattage requirement in the glove box can be classified into two
categories, first to raise the temperature of the incoming air and secondly for
raising the temperature of the material in the composting vessels.

Air wattage requirement

As per Omega Engineering catalogue for air process heater, power
requirement can be calculated as:

Watts = SCFM x 'T / 3

SCFM = Standard cubic feet per minute

"T = Temperature rise in degrees F from the inlet to the outlet
Exit air from AHP series heaters can reach temperatures upto 540°C (1000°F). In
our experiment, exit air temperatures of around 250°C were recorded. Therefore,
air wattage requirement was calculated based on the temperature rise from room
temperature of 21°C to 250°C. As mentioned in section 5.2, air-flow to each
glove box was recorded in the range of 50 - 60 SCFH, thus air-flow of 1 SCFM
was assumed for this calculation.

'1' = 482°F — 69.8°F = 412.2°F

Wattage,“r = 1 SCFM x 412.2°F / 3 = 137 Watts
Composting vessels wattage requirement

Each composting vessel in the glove box consisted of 600 grams of dry
solids of compost with or without 100 grams of dry solids of test substance and
the dry solids of the overall sample adjusted to approximately 50%, as per ASTM

standard D 5338. Each glove box had capacity to hold 15, 2 Liter Erlenmeyer

286

Flasks, which were used as composting vessels. For simplifying the calculations
each composting vessels was assumed to contain 600 grams of compost plus
100 grams of extruded-injection molded Poly(3-hydroxybutyric acid) (PHB)
dogbone samples.

Therefore, wattage requirement for the composting vessels was calculated
by determining the energy required to raise the temperature of a mix of compost,
test substance (assumed all the samples as PHB) and water to 58°C.

To determine the energy required (Q), specific heat capacity (C,,) of the
materials should be known and calculated as:

Q = m x Cp x AT;

where:

Q = energy in Joules

m = weight of substance in grams

CD = specific heat capacity of the substance in J/g/K

AT = temperature rise in °C / °K
Mears et al.1 have calculated specific heat capacity for the compost, obtained
from a mix of swine wastes and straw. For a 35-day mature dry compost (with
zero% moisture), specific heat is reported as 0.1551 cal/g/°C. But in the current
study this value was not used because Mears et al. reported that swine waste
used for the study contained significant amount of inorganic materials such as
bones, glass etc., thus modifying the actual specific heat numbers. In addition the
dry compost specific heat value is extrapolated from calculations made at higher

moisture content, therefore increasing the error of such a value.

287

In the current study yard-waste based compost was used and screened to
remove any large inert items (glass, stone, wood, etc.). As mentioned by
Narayanz, compost provides organic matter in eroded soils, acting well as a
substitute for organic carbon in soils. Therefore, specific heat of soil organic
matter reported as 0.46 cal/g/°C by Farouki3 was used to estimate the energy
requirement for compost.

Poley et al.4 have calculated specific heat capacity of PHB films measured
by temperature rise method under continuous white-light illumination and
reported as 3.98 :I: 0.20 (J/cm3/K). Using the density5 value of 1.25 g/cm3,
specific heat capacity was converted to mass units as 3.184 1: 0.16 (J/g/K).

Wattage requirement for composting vessels was estimated based on
heating the vessels in an hour.

Wattagecomposung = n x [ (Woompost x Cmmpos.) + (WpHB x CppHa) + (WWfir x 0pm“) ] x AT / t;

n = number of composting vessels = 15

Woompost = weight of compost per vessel = 600 grams

WpHB = weight of PHB per vessel = 100 grams

Wwate, = weight of water per vessel = 700 grams

Cpmmpost = specific heat of compost = 0.46 cal/g/°C = 1.925 J/g/K

Cp-pHB = specific heat of PHB = 3.184 J/g/K

Cp-wate, = specific heat of water6 = 4.1774 J/g/K

AT = temperature rise in °C / °K = 37

t = heating time for the vessels = 1 hour = 3600 sec

Thus wattage requirement for composting vessels was calculated as:

288

 

wattagecomposting = 678 watts;
And total wattage requirement for each glove box was calculated as:

wattagetotal = wattageair + wattagecomposting = 815 watts.

289

 

References:

 

‘ Mears OR. at al., Thermal and physical properties of compost. Energy, agriculture, and
waste management: proceedings of the 1975 Cornell Agricultural Waste Management
Conference.

2 R. Narayan, Drivers for biodegradable/compostable plastics & role of composting in
waste management & sustainable agriculture. Bioprocessing of Solid Waste & Sludge,

Vol. 1, 2001.
3 Omar T. Farouki, Thermal Properties of Soils. Series on Rock and Soil Mechanics, Vol.

11(1986)

4 Poley, L.H. et al., Photothermal Methods and Atomic Force Microscopy Images Applied
to the Study of Poly(3-Hydroxybutyrate) and Poly(3-Hydroxybutyrate-co-3-
Hydroxyvalerate) Dense Membranes. Journal of Applied Polymer Science, Vol. 97
2005)

g Material Safety Data Sheet PHB, Biomer, www.biomer.de

6 Perry's Chemical Engineers’ Handbook (7th Edition), 1997

290

Appendix 5.6.3: Sample Calculation for carbon dioxide evolution on
mass basis

Sample calculation for converting concentration reading for Compost

Blank sample is provided below. This calculation is for the first reading after

starting the experiment.
Sample: Compost Blank-7A
Date of reading: 12/16/2004
Gas Chromatograph Reading:
Retention time (minutes): 1.166
Area: 22.4780
% age 002: 1.972

Note; As discussed in section 5.2 (Experimental Set-up and
Procedure) of my thesis, two different C02 calibration curves were
developed: one for a C02 concentration range of 20% - 0.2% and another
of 3% - 0.2% range. In the reading above, calibration curve with 3% -
0.2% range was used for greater accuracy in determining CO;
concentration.

Flow rate of the exhaust gas: 80 mL/min = 1.33 mL/sec

V002 = sample volume of C02 = (% age C02) x (Flow rate of the exhaust

gas) = 0.0263 ml:
sec

Using ideal gas law equation and room temperature & pressure

conditions, mass of C02 was calculated as follows:

291

P X V002 X M002

m = , where
002 R x T

 

P = 1 atmosphere
T = 298 Kelvin
R = gas constant = 82.057 (mL.atm/moles.K)

M002 = molecular weight of CO2 = 44

Therefore mass of C02 (mCO2 ) was calculated as: 4.731 x E — 05 i

sec
Start up date and time: 12/15/2004, 5:00 pm
Flowrate adjustment date and time: 12/16/2004, 6:30 am
Net days (No): 0.56

Thus grams of 002 =4.731xE-05 9 x3600§5§x24£§x056 day

reading sec

 

 

= 2.299 9_
reading

 

Converting grams of CO? to carbon, C9 = 2.299—£7— x12
read ng reading 44

g of C
reading

 

C9 = 0.627

Dry weight of the compost blank sample = 599.95 9

g of C02 Ireadlng = 0.0038.
g of sample

 

Therefore,

292

 

 

 

   

llllllllllllllflllllllllllllllll“illilllllllllll
3 1293 02845 7814